Novel & Intelligent Digital Systems: Proceedings of the 2nd International Conference (NiDS 2022) (Lecture Notes in Networks and Systems) 3031176006, 9783031176005

Citation preview

Lecture Notes in Networks and Systems 556

Akrivi Krouska Christos Troussas Jaime Caro   Editors

Novel & Intelligent Digital Systems: Proceedings of the 2nd International Conference (NiDS 2022)

Lecture Notes in Networks and Systems Volume 556

Series Editor Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland Advisory Editors Fernando Gomide, Department of Computer Engineering and Automation—DCA, School of Electrical and Computer Engineering—FEEC, University of Campinas— UNICAMP, São Paulo, Brazil Okyay Kaynak, Department of Electrical and Electronic Engineering, Bogazici University, Istanbul, Turkey Derong Liu, Department of Electrical and Computer Engineering, University of Illinois at Chicago, Chicago, USA Institute of Automation, Chinese Academy of Sciences, Beijing, China Witold Pedrycz, Department of Electrical and Computer Engineering, University of Alberta, Alberta, Canada Systems Research Institute, Polish Academy of Sciences, Warsaw, Poland Marios M. Polycarpou, Department of Electrical and Computer Engineering, KIOS Research Center for Intelligent Systems and Networks, University of Cyprus, Nicosia, Cyprus Imre J. Rudas, Óbuda University, Budapest, Hungary Jun Wang, Department of Computer Science, City University of Hong Kong, Kowloon, Hong Kong

The series “Lecture Notes in Networks and Systems” publishes the latest developments in Networks and Systems—quickly, informally and with high quality. Original research reported in proceedings and post-proceedings represents the core of LNNS. Volumes published in LNNS embrace all aspects and subfields of, as well as new challenges in, Networks and Systems. The series contains proceedings and edited volumes in systems and networks, spanning the areas of Cyber-Physical Systems, Autonomous Systems, Sensor Networks, Control Systems, Energy Systems, Automotive Systems, Biological Systems, Vehicular Networking and Connected Vehicles, Aerospace Systems, Automation, Manufacturing, Smart Grids, Nonlinear Systems, Power Systems, Robotics, Social Systems, Economic Systems and other. Of particular value to both the contributors and the readership are the short publication timeframe and the world-wide distribution and exposure which enable both a wide and rapid dissemination of research output. The series covers the theory, applications, and perspectives on the state of the art and future developments relevant to systems and networks, decision making, control, complex processes and related areas, as embedded in the fields of interdisciplinary and applied sciences, engineering, computer science, physics, economics, social, and life sciences, as well as the paradigms and methodologies behind them. Indexed by SCOPUS, INSPEC, WTI Frankfurt eG, zbMATH, SCImago. All books published in the series are submitted for consideration in Web of Science. For proposals from Asia please contact Aninda Bose ([email protected]).

More information about this series at https://link.springer.com/bookseries/15179

Akrivi Krouska Christos Troussas Jaime Caro •

•

Editors

Novel & Intelligent Digital Systems: Proceedings of the 2nd International Conference (NiDS 2022)

123

Editors Akrivi Krouska University of West Attica Aigaleo, Greece

Christos Troussas University of West Attica Aigaleo, Greece

Jaime Caro College of Engineering University of the Philippines Diliman, Philippines

ISSN 2367-3370 ISSN 2367-3389 (electronic) Lecture Notes in Networks and Systems ISBN 978-3-031-17600-5 ISBN 978-3-031-17601-2 (eBook) https://doi.org/10.1007/978-3-031-17601-2 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Preface

The 2nd International Conference on Novel & Intelligent Digital Systems (NiDS2022) was held in Athens, Greece, from September 29 to 30, 2022. The conference was implemented virtually due to COVID-19, under the auspices of the Institute of Intelligent Systems (IIS). The Hosting Institution of NiDS2022 was the University of West Attica (Greece). NiDS lays special emphasis on the novelties of intelligent systems and on the interdisciplinary research for enabling, supporting and promoting artificial intelligence (AI) in software development. It promotes high-quality research, creating a forum for exploration of challenges and novel advancements in AI. It triggers an exchange of ideas in this field, reinforcing and expanding the network of researchers, academics and market representatives. NiDS addresses experts/researchers and scholars in the fields of artificial and computational intelligence in systems, as well as computer science in general, enabling them to learn more about pertinent fields, which are strongly related and mutually complementary. Topics within the scope of NiDS series include but are not limited to: Adaptive systems Affective computing Augmented reality Big data Bioinformatics Cloud computing Cognitive systems Collaborative learning Cybersecurity Data analytics Data mining and knowledge extraction Decision-making systems Deep learning Digital marketing

v

vi

Preface

Digital technology Distance learning E-commerce Educational data mining E-learning Environmental informatics Expert systems Fuzzy systems Genetic algorithm applications Human–machine interaction Information retrieval Intelligent information systems Intelligent modeling Machine learning Medical informatics Mobile computing Multi-agent systems Natural language processing Neural networks Pattern recognition Personalized systems and services Pervasive multimedia systems Recommender systems Reinforcement learning Semantic web applications Sentiment analysis Serious gaming Smart cities Smart grid Social media applications Social network analytics Text mining Ubiquitous computing User modeling Virtual reality Web intelligence The call for scientific papers solicited work presenting substantive new research results in using advanced computer technologies and interdisciplinary research for enabling, supporting and enhancing intelligent systems. The international program committee consisted of 52 leading members of the intelligent systems community, as well as highly promising younger researchers. The conference (General) chairs were Cleo Sgouropoulou and Ioannis Voyatzis from University of West Attica (Greece), whereas the program committee chairs were Akrivi Krouska and Christos Troussas from University of West Attica and

Preface

vii

Jaime Caro from University of the Philippines. The keynote speaker of NiDS 2022 was Prof. Mirjana Ivanovic, a full professor at Faculty of Sciences, University of Novi Sad, Serbia. The title of her speech was “How Artificial Intelligent Approaches Support Medical Decisions and Patients’ Wellbeing”. Scientific papers were reviewed rigorously by two to three reviewers (one of which was senior) through a double-blind process, thus reflecting our commitment to make NiDS a top-flight, selective and high-quality conference. We believe that the chosen full papers describe some very significant research, while the short papers present some very interesting new ideas. In the review process, the reviewers’ evaluations were generally respected. The management of the review process and the preparation of the proceedings were handled through EasyChair. We would like to thank all those who have contributed to the conference, the authors, the program committee members and the organization committee with its chair, Kitty Panourgia, as well as the Institute of Intelligent Systems. Akrivi Krouska Christos Troussas Jaime Caro

Committees

Conference Committee General Conference Chairs Cleo Sgouropoulou Ioannis Voyatzis

University of West Attica, Greece University of West Attica, Greece

Honorary Chair Claude Frasson

University of Montreal, Canada

Program Committee Chairs Akrivi Krouska Christos Troussas Jaime Caro

University of West Attica, Greece University of West Attica, Greece University of the Philippines

Program Advising Chair Peter Hajek

University of Pardubice, Czech Republic

Workshop and Tutorial Chair Athanasios Voulodimos

National Technical University of Athens, Greece

Doctoral Consortium Chair Dimitris Sotiros

Wrocław University of Science and Technology, Poland

ix

x

Committees

Organization Chair Kitty Panourgia

Neoanalysis Ltd, Greece

Publicity Chairs Karima Boussaha Oleg Sychev Shahzad Ashraf

University of Oum El Bouaghi, Algeria Volgograd State Technical University, Russia Hohai University, China

The conference is held under the auspices of the Institute of Intelligent Systems.

Program Committee Ali Abd Almisreb Shahzad Ashraf Abul K. M. Azad Costin Badica Maumita Bhattacharya Siddhartha Bhattacharyya Karima Boussaha Ivo Bukovsky George Caridakis Jaime Caro Adriana Coroiu Athanasios Daradoumis Samia Drissi Eduard Edelhauser Kurt Junshean Espinosa Andreas Floros Foteini Fotopoulou Claude Frasson Peter Hajek Layla Hasan Nantia Iakovidou Katerina Kabassi Athanasios Kakarountas Yasushi Kambayashi Achilles Kameas

IUS, Bosnia and Herzegovina Hohai University, China Northern Illinois University, USA University of Craiova, Romania Charles Sturt University, Australia Christ University, India University of Oum El Bouaghi, Algeria CTU, Czech Republic University of the Aegean, Greece University of the Philippines, Philippines Babeș-Bolyai University, Romania University of the Aegean, Greece University of Souk Ahras, Algeria University of Petrosani, Romania University of Manchester, UK Ionian University, Greece University of Patras, Greece University of Montreal, Canada University of Pardubice, Czech Republic University of Technology Malaysia King’s College London, UK Ionian University, Greece University of Thessaly, Greece Nippon Institute of Technology, Japan Hellenic Open University, Greece

Committees

Zoe Kanetaki George Kolezas Petia Koprinkova-Hristova Akrivi Krouska Florin Leon George Magoulas Phivos Mylonas Vasileios Nittas Vera Novikova Stavros Ntalampiras Lanndon A. Ocampo Kyparisia Papanikolaou Nikolaos Polatidis Spyros Polykalas Theodosios Sapounidis Filippo Sciarrone Cleo Sgouropoulou Allan Sioson Dimitris Sotiros Antonio Staiano George Styliaras Christos Troussas Panagiotis Vlamos Athanasios Voulodimos Ioannis Voyiatzis Davide Zambrano

xi

University of West Attica, Greece NTUA, Greece Bulgarian Academy of Sciences, Bulgaria University of West Attica, Greece Technical University of Iasi, Romania Birkbeck College, University of London, UK Ionian University, Greece University of Zurich, Switzerland Tomsk Polytechnic University, Russia University of Milan, Italy Cebu Technological University, Philippines ASPETE, Greece University of Brighton, UK Ionian University, Greece International Hellenic University, Greece Roma Tre University, Italy University of West Attica, Greece Ateneo de Naga University, Philippines WUST, Poland University of Naples Parthenope, Italy University of Patras, Greece University of West Attica, Greece Ionian University, Greece NTUA, Greece University of West Attica, Greece EPFL, Switzerland

Contents

Approaches in Adaptive Learning Cross-Cutting Visual Support of Decision Making for Forming Personalized Learning Spaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Viktor Uglev and Tatiana Gavrilova Personalized Learning in an Intelligent Educational System . . . . . . . . . Valentina Terzieva, Tatyana Ivanova, and Katia Todorova Electronic-Service Learning to Sustain Instruction with Civic Engagement During the COVID-19 Pandemic . . . . . . . . . . . . . . . . . . . . Aurelio Vilbar Evaluating E-Learning Process on Virtual Classroom Systems Using an ISO-Based Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nicholas Coulianos, Athanasia Sapalidou, Akrivi Krouska, Christos Troussas, and Cleo Sgouropoulou

3 13

24

33

SERVE as Instructional Design for Low-Connectivity Online Self-directed Modules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jeraline Gumalal and Aurelio Vilbar

46

Extended Technology Acceptance Models for Digital Learning: Review of External Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Akrivi Krouska, Christos Troussas, and Cleo Sgouropoulou

52

Extended Reality and Games Designing a VR Application for Typhoon Preparedness Training in a Classroom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Barbara P. David, Neil Jherome L. Hernandez, Kim Farhant S. Palao, Richelle Ann B. Juayong, and Jaime D. L. Caro

67

xiii

xiv

Contents

Virtual Reality in Education: Reviewing Different Technological Approaches and Their Implementations . . . . . . . . . . . . . . . . . . . . . . . . . Andreas Marougkas, Christos Troussas, Akrivi Krouska, and Cleo Sgouropoulou Ready to Play - A Comparison of Four Educational Maze Games . . . . . Elena Paunova-Hubenova, Yavor Dankov, Valentina Terzieva, Dessislava Vassileva, Boyan Bontchev, and Albena Antonova Employing FFNN and Learning Styles to Improve Knowledge Acquisition in Educational Digital Games . . . . . . . . . . . . . . . . . . . . . . . Christos Troussas, Akrivi Krouska, and Cleo Sgouropoulou

77

84

95

Effectiveness of Open-Source Solutions for Limited Scale Interventions Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Ioannis Kavouras, Emmanuel Sardis, Eftychios Protopapadakis, and Anastasios Doulamis Modeling the Knowledge of Users in an Augmented Reality-Based Learning Environment Using Fuzzy Logic . . . . . . . . . . . . . . . . . . . . . . . 113 Christos Papakostas, Christos Troussas, Akrivi Krouska, and Cleo Sgouropoulou Evaluating the Feasibility of Fast Game Development Using Open Source Tools and AI Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 Ioannis Kavouras, Ioannis Rallis, Anastasios Doulamis, and Nikolaos Doulamis A 2D Platform Game Offering Personalized Guidance to Players to Promote Environmental Awareness . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Argyris Sideris, Christos Troussas, Akrivi Krouska, and Cleo Sgouropoulou Health and Earth Science Application Design for a Virtual Reality Therapy Game for Patients with Behavioral and Psychological Symptoms of Dementia . . . . . . . . . . 149 Veeda Michelle M. Anlacan, Roland Dominic G. Jamora, Angelo Cedric F. Pangilinan, Isabel Teresa O. Salido, Maria Evelyn V. Jacinto, Michael L. Tee, Maria Eliza R. Aguila, Cherica A. Tee, and Jaime D. L. Caro Development of Methods and Models for Assessing Spine Curvature Based on Antilatency Motion Capture System . . . . . . . . . . . . . . . . . . . . 161 Anastasia Romanovna Donsckaia, Yulia Alexandrovna Orlova, Stanislav Vladislavovich Stepanov, Dmitry Romanovich Cherkashin, and Viktor Viktorovich Noskin

Contents

xv

Application Design of a Virtual Reality Therapy Game for Patients with Cerebral Palsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Maria Eliza R. Aguila, Cherica A. Tee, Josiah Cyrus R. Boque, Isabel Teresa O. Salido, Maria Evelyn V. Jacinto, Michael L. Tee, Veeda Michelle M. Anlacan, Roland Dominic G. Jamora, and Jaime D. L. Caro How Artificial Intelligent Approaches Support Medical Decisions and Patients’ Wellbeing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Mirjana Ivanovic A Landslide Model Using a 3D Ultradiscrete Burgers’ Equation . . . . . . 190 Lee Javellana Gamified Upper-Limb Rehabilitation Program for Elderly Participants Using a Real-Time Motion Tracking System . . . . . . . . . . . 200 Vitus Murdock F. Acabado, Gianna Pauline B. Burgos, Jaime D. L Caro, Richelle Ann B. Juayong, and Maria Eliza Ruiz Aguila An Analysis of Mental Workload Involved in Piloting Tasks . . . . . . . . . 211 Maryam Ghaderi, Hamdi Ben Abdessalem, and Claude Frasson Architecture of the Android Application for Monitoring Person’s Condition Based on Data Readings from Sensors of Smart Watches and Mobile Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Vadim Viktorovich Gilka, Yuri Alexandrovich Kachanov, and Agnessa Sergeevna Kuznetsova Design Strategies on Virtual Reality for Cognitive Monitoring of Older Persons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232 Angelo Cedric F. Panganiban, Jaime D. L. Caro, Richelle Ann B. Juayong, and Veeda Michelle M. Anlacan Information Systems and Science Predictive and Prescriptive Business Process Monitoring with Reinforcement Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 Silvester Kotsias, Athanasios Kerasiotis, Alexandros Bousdekis, Georgia Theodoropoulou, and Georgios Miaoulis Unified Graphic Visualization of Activity (UGVA) Method . . . . . . . . . . 255 Viktor Uglev Disk Space Consumption by Triple Storage Systems . . . . . . . . . . . . . . . 266 Artem Prokudin, Mikhail Denisov, and Oleg Sychev The Relationship of Disability, New Technologies, and ‘Smart Packaging’: The Greek Experience . . . . . . . . . . . . . . . . . . . 276 Maria Poli and Konstantinos Malagas

xvi

Contents

Modern Approaches for Concepts and Relations Extraction for Ontology Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 Alexander Katyshev and Anton Anikin Attentional Tasks Model: A Focus Group Approach . . . . . . . . . . . . . . . 297 Maryam Ghaderi, Marc-Antoine Courtemanche, Hamdi Ben Abdessalem, Roger Nkambou, and Claude Frasson From Threads to Textiles: Building an Ontology for the Indigenous Fabrics of the Ifugao . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308 Herbert Gerard T. Villafranca, Jaime dL. Caro, Romanlito S. Austria, and Analyn Salvador-Amores Surveying Search Terms for COVID-19 Disease Surveillance . . . . . . . . 318 Adrian Galido and Jerina Jean Ecleo Development of Models and Methods for Building a Psychological Portrait of a Person Based on Information from Social Networks . . . . . 328 Vladimir A. Litvinenko, Roman V. Titov, Alexander V. Zubkov, Yulia A. Orlova, and Yana V. Kulikova Data Mining and Machine Learning A Deep Convolutional Neural Network for Skin Rashes Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 Jannie Fleur V. Oraño, Francis Rey F. Padao, and Rhoderick D. Malangsa Evaluating YOLO Transferability Limitation for Road Infrastructures Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349 Iason Katsamenis, Agapi Davradou, Eleni Eirini Karolou, Eftychios Protopapadakis, Anastasios Doulamis, Nikolaos Doulamis, and Dimitris Kalogeras Developing Novel Learning Spaces Through Social Media Channels for Sustainable CAD Engineering Education . . . . . . . . . . . . . . . . . . . . . 359 Zoe Kanetaki, Constantinos Stergiou, Christos Troussas, and Cleo Sgouropoulou Greek Patent Classification Using Deep Learning . . . . . . . . . . . . . . . . . 372 Ioannis Pontikis, Stratos Koutivas, Panagiotis Kasnesis, Alexandria Filippou, and Dimitris Stafylas TraCon: A Novel Dataset for Real-Time Traffic Cones Detection Using Deep Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382 Iason Katsamenis, Eleni Eirini Karolou, Agapi Davradou, Eftychios Protopapadakis, Anastasios Doulamis, Nikolaos Doulamis, and Dimitris Kalogeras

Contents

xvii

Machine Learning Methods for Modeling Dengue Incidence in Local Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392 Jozelle C. Addawe, Jaime D. L. Caro, and Richelle Ann B. Juayong Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401

Approaches in Adaptive Learning

Cross-Cutting Visual Support of Decision Making for Forming Personalized Learning Spaces Viktor Uglev1(B)

and Tatiana Gavrilova2

1 Siberian Federal University, Krasnoyarsk, Russia

[email protected] 2 St. Petersburg University, St. Petersburg, Russia

Abstract. The paper describes an approach to support methodological decisionmaking processes and explain them in the Intelligent Tutoring Systems (ITS), relying on cognitive visualization tools. The importance of XAI problems in the framework of automatic analysis of hypotheses concerning the interpretation of the learning situation is noted. In order to combine the structural and functional approaches within a single logic of analysis, various graphic notations are considered. Vertical (for different scopes of learning situation coverage) and horizontal (for different aspects of analysis) transitions between decision-making levels are shown for the proposed decision-making model. A combination of the method of Cognitive Maps of Knowledge Diagnosis (CMKD, structural aspect) and the method of Unified Graphic Visualization of Activity (UGVA, functional aspect) is taken as the basis. The steps of the process of superimposing data from the digital educational footprint on invariant graphic notations are described. The importance of ensuring the isomorphism of displaying the parameters of the learning situation when switching between maps and images is noted. Visualizations of CMKDs and anthropomorphic UGVA images are presented to illustrate examples from the real learning process. The advantages of the cross-cutting visual support for both the decision-making process and the synthesis of explanatory texts are shown. In conclusion, the limitations of the approach are pointed out, and the directions for further research are outlined. Keyword: Decision making · Intelligent Tutoring Systems · Cognitive visualization · Decision explanation · Personalized learning space

1 Introduction and Related Works The modern e-learning space is shaped largely by automated systems with elements of artificial intelligence, called Intelligent Tutoring Systems (ITS). Such systems shall not only analyze the learning environment and make decisions, but also format the feedback collected from human users (mainly students and teachers). This is important due to the fact that the human tends not to trust the decisions of systems built on the basis of artificial intelligence, and needs the possibility to verify the adequacy of system reactions (the concept of XAI [2]). © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 3–12, 2023. https://doi.org/10.1007/978-3-031-17601-2_1

4

V. Uglev and T. Gavrilova

The data used by the ITS refer to the electronic course, to the parameters and actions of the student (digital footprint), as well as to the knowledge base entered into the training system. Taking the system approach [7] as the basis, the analysis procedure shall be carried out simultaneously in the functional and structural aspects. At the same time the processing of the learning situation shall cover the following types of parameters if possible (see [12] for details): – – – –

decision-making scope (PM : nano, micro, meso, macro and meta level, see Fig. 1a); levels of decision making (PL : operational, tactical, strategic); aspects of consideration (PK : subject, competence, normative, target); temporal dynamics (PN : retrospective data, current indicators, projected/planned indicators).

The links between the structural components of the learning material in different levels of education, organized as a set of academic disciplines, are grouped not only by semesters (see the example of the specialty tree in Fig. 1b), but also correspond to different major functional aspects (major groups of key skills, identified on the basis of competency model, are coded by color). Notably, the links in the hierarchy are nonlinear and do not allow to unambiguously correlate discipline profiles with the groups of competencies or key skills (it is a many-to-many relationship). On this basis, the local conclusions of ITS, which affect the current learning situation, shall either take into account the ambiguity of these links, or shall be guided by the strategic goals of using the learning material. On the one hand, this approach complicates the decision-making process and, on the other hand, it increases its flexibility in terms of actions and synthesis of explanatory decisions.

Fig. 1. Scopes of decision-making in relation to the learning material (a) and structural hierarchy of the learning material from the macro level (b, the example of the master’s degree program in Informatics and Computer Science)

Interpretation of ITS decisions in a human-accessible form, as a rule, is reduced to the synthesis of a text in the natural language form (text of a dialogue or its voicing by a virtual assistant) or in the form of a graphic image. If the subject of explanation is simple, it is well perceived by ear. Otherwise, a large volume of text and/or the formation of explanatory images is required [6, 9]. Educational programs, e-courses and even a digital

Cross-Cutting Visual Support of Decision Making

5

educational footprint of a student are all quite complex both structurally and functionally. Therefore, accessible reasoning of ITS actions shall provide dialectical unity (according to Plato) of form and content. There are different approaches to the organization of the description structure and logic of the analysis of the learning situation, which are based on: – the ontology of the subject area [8, 13]; – curricula and standards of training programs [14]; – models of particular courses [3]. The learning process, like any process, needs to be analyzed in dynamics. This requires taking into account the dynamics of events (past), current parameters (present), and projected/planned parameters (future). This formulation of the problem corresponds to Anokhin’s process of afferent synthesis [1]. Separating and simultaneously considering structural and functional specificity of relations between entities involved in the learning process, we can compare graphic notations used by developers of modern ITS (see Table 1). Table 1. Characteristics of graphic notations used for learning material description Notation

Complexity of representation

Scalability

Consideration of dynamics

Structure/function transition

Graphs

–

–

V

–

Hierarchy graphs

V

V

–

–

Dashboards

V

–

–

–

Mental and conceptual maps

V

~V

–

–

(CMKD)

V

V

~V

~V

Semantic networks and ontologies

V

~V

~V

~V

Factor space maps

V

V

–

–

Clusters

V

–

–

–

Pictographs (Chernoff’s faces)

V

–

–

–

UGVA

V

V

~V

~V

As we can see from the table, it is difficult to take into account the structural and functional aspects systematically using the above notations, because they rely on different types of learning situation parameters. It follows from the last column that the graphic notation alone is not enough to solve this problem. Besides, using ontologies requires a total detailed description of a subject domain, which is difficult in real education process conditions. To overcome the difficulty, it is necessary to operate with a minimum set of

6

V. Uglev and T. Gavrilova

notations, which will allow ITS isomorphic display of structural and functional aspects of the learning situation in an automatic mode for an arbitrary combination of values of the analyzed parameters from P. The purpose of this paper is to briefly describe the idea of the cross-cutting approach to graphic support of personalized decision-making processes in ITS using visual notations. For this purpose, we will describe the basic model of data organization for the cross-cutting analysis of a learning situation. We will describe the process of isomorphic display of structural representation by means of Cognitive Maps of Knowledge Diagnosis (CMKD) and functional representation by means of the Unified Graphic Visualization of Activity (UGVA) method. We will also give a scheme of the cross-cutting analysis using graphic notations and give some illustrative examples.

2 Method 2.1 Basic Decision-Making Model Formalizing a decision-making process involves determining the input data to be analyzed. Let us combine the following data: – a model of learning material (from individual tests/assignments to a curriculum with a structural description and linkage to activity outcomes) supplemented with semantic links (S, corresponds to Fig. 1b) and methodological goals (G1 ); – digital educational footprint of the student as part of the student model, including student’s declared preferences and goals (G2 ), as well as data on personal learning trajectory in the form of punctures of his/her actions (R); – a student-initiated or ITS-initiated event (E) that requires not only decision making but also dialogue with the student. Then the decision-making model as an afferent synthesis mechanism [1] implemented by an intelligent ITS solver will correspond to (1). F( γ , E, R|S) → (P, G) → ,

(1)

where y is the chosen decision for the current situation, and q is the resulting explanation (including graphic elements) corresponding to the situation P at time γ. The parameter a depends on both the specificity of the learning situation and the P analysis parameters, determined by the mechanism of expert systems, and in the synthesis q it is expressed by the choice of emphasis in favor of the structural or functional aspect of the visual support. To explain the main stages of transition from input data to decisions, let us refer to the scheme shown in Fig. 2. Through E (γ, q) we denote the event, the transition between P components (relative to the sets M, L, K and N), initiated by the user when developing the explanatory dialogue.

Cross-Cutting Visual Support of Decision Making

7

Fig. 2. Scheme of implementation of the cross-cutting analysis of the learning situation using cognitive data visualization tools.

Personalization of the decision and explanations is achieved both by taking into account the digital educational footprint and by operating the target indicators G in the context of the most relevant combinations of display conditions in P. 2.2 Structural and Functional Visualization The purpose of graphic support of ITS decisions and responses is to induce the student to consciously accept (trust) the responses of the system regarding the methodological side of the educational process. The general idea of applying the means of cognitive visualization lies in the fact that the information from an individual digital educational footprint is superimposed on the invariant substrate of the structural or functional space of analyzed objects in one of the scopes from M so that it can be easily interpreted regarding the parameters L, K and N in accordance with the dialogue (E(γ1 ) → E (γ2 ) → …). The display of the current values of the key indicators is coded through color (the heat map), and the secondary indicators are coded through other parameters (thickness of lines, shape of elements, inscription parameters, etc.). The possibility to combine current, retrospective and target indicators within a single image allows to apply mapping (1) with comparable detailing with respect to both structural and functional aspects (isomorphism effect). As a basis for structural representation we use the CMKD method [11], which forms a map of the learning material for all scope levels from micro to macro (see Fig. 1) as a “small world” model with semantic links. Figure 3a shows an example of a map for the academic semester of the Informatics and Computer Science program (nodes contain academic modules and edges contain semantic links) and superimposition of the data from the individual digital educational footprint in the subject (Fig. 3b) and competence aspects (Fig. 3c for competence UK-1 “Can critically analyze problem situations based on the system approach and develop the action strategy”). A detailed description of data superimposition and its interpretation is given in [11]. Here we will note, that the student will see that variant of a map (or its fragment) which corresponds to the most significant accent in an explanation of the decision taken by ITS with possibility to pass to E‘ in other aspects (including change of a scope).

8

V. Uglev and T. Gavrilova

Fig. 3. Example of the CMKD for academic disciplines of one semester (a), superimposing the data of one of the students on the map in the subject aspect (b, grades in points) and competency aspect (c, detailed assessment of the level of mastery of one of the competencies)

The basis for functional representation will be the UGVA notation (type λ5Bit), which forms the image of the activity-based description of the learning process for all levels of scope (see Fig. 1) as an anthropomorphic image [10], developing the idea of Trizkin modulor [4]. Figure 4 a shows an example of the image for the curriculum of the Informatics and Computer Science program (linear dimensions of each block are the value of the contribution of each group of key skills in the core and selective development for the current stage of education in accordance with the colors from Fig. 1b). When the data from the digital educational footprint of our chosen student is superimposed on the image, we get images in the subject (Fig. 4b) and competence (Fig. 4c) aspects. Three legs correspond to the arms and body (horizontal axis of symmetry), relative to similar indicators for the previous level of education. A detailed description of the data superimposition and its interpretation is given in [10]. Similar to the CMKD, the image is interactive and allows images to be obtained through dialogue in the context of different accents from P.

Fig. 4. An example of an image in UGVA notation for the curriculum of Informatics and Computer Science program (a), superimposition of one of the students’ data on the image in the subject (b, grades in points) and competence aspects (c, supra-subject assessment of the mastery of a set of competencies).

Cross-Cutting Visual Support of Decision Making

9

Both graphic notations are synthesized automatically by the ITS intellectual core. 2.3 Cross-Cutting Approach The essence of the cross-cutting approach is that the ITS intellectual solver at a point in time γ at the event E enters the space defined by the P analysis parameters for the actual value M. The invariant graphic images of CMKD and UGVA methods are specified using the current values of L, K and N, extracting and grouping the data from the digital educational footprint necessary for the decision making and explanation (metric concentration). This makes it possible not only to automatically switch (redraw) maps and images between aspects of the displayed P (see Fig. 5a), but also to change the scope according to the needs of the analysis process (see Fig. 5b). Since the visualizations are interactive, the user can get both explanations and details, as well as go beyond the current “focus” of consideration to understand why and wherefores the current decisions were made in the ITS. Flexible transition between visualizations is carried out due to free switching both vertically (on a scope from M), and horizontally (on other components from P). Isomorphism of displaying the digital footprint parameters in structural and functional aspects between notations can vary with respect to the role of the recipient person (student, teacher, methodologist). By forming and displaying the map/image to the user, a dialogue text is automatically formed. It is accompanied by a set of clarification questions of methodological character with parameters that the person can choose. The route of the dialogue is organized both for clarification of the prerequisites (inverse solver display strategy) and for clarification of the significance (direct display strategy). At the same time, CMKDs and UGVA images can be displayed only partially, increasing the concentration of the most significant data and emphasizing them to the user (for example, by additional color highlighting). Forming an appropriate variant of the query to ITS the user proceeds to a new explanation accompanied by an appropriate visualization (E → E ).

Fig. 5. Variability of horizontal (a) and vertical (b) parameters of image synthesis in the crosscutting approach to supporting the processes of ITS decision making and explanation

10

V. Uglev and T. Gavrilova

3 Case Study and Discussion Let’s consider examples of visualization in relation to the real educational process. As a basic program we choose the master’s degree program in Informatics and Computer Science at Siberian Federal University. For the current group of students (2021) the specialty tree will look as shown in Fig. 1b (macro and meso levels [12]). Each discipline Dj corresponds to a set of learning topics (nodes t i in the graph), divided into didactic units supplemented with the test material. When forming a personal learning trajectory, the individual elements of the learning material (didactic units, topics and even entire training courses) may not be included in the training program (reflected on the outer contour of the circle in the CMKD). In the process of training on the basis of ITS the group of students solves tasks and tests, which are evaluated in terms of the residual knowledge (subject aspect) and the level of competence mastery. This data complements each student’s individual digital educational footprint and allows for a date-specific inquiry of grades and a competency profile in the ITS. Combining the current indicators, results of the previous tests and student preferences (taken from questionnaires), the ITS solver makes decisions (grading, recommendations, predicting learning dynamics, motivation to intensify work with the learning material, etc.), activating a chain of steps shown in Fig. 2. Let us choose one of the students as an object to illustrate the results of structural representation. Figure 3b shows the CMKD for the learning material from the meso level (the third academic semester), reflecting the student’s subject grades. It can be seen that there is a lack of mastery of the learning material based on the results of the knowledge assessments of subject topics t5, t7, t14, and t15. Figure 3c shows the map corresponding to the data on the competence UK-1 from its current supra-subject competence profile. It shows that for the successful mastery of this competence it is necessary to improve the understanding of the material from the subject topics t7 and t12. The data from these two maps can be combined to make decisions about the synthesis of the prompt within the discipline D3.2 : the problem of accounting for interdisciplinary links (t7 is influenced by t12, which depends on t14) clearly emerges. If the student, who has been recommended to turn first to t14 from D3.5 and to t5, asks the question “Why t14 and t12, and not t7?”, he gets the semantic links marked on the map shown in Fig. 3b (red arrows) and also the comments on his earlier preferences for mastering UK-1 according to his questionnaire (in the dialogue it is possible to get the map shown in Fig. 3c based on competencies and to display a red arrow t7-t12). It is possible to expand the depth of analysis by displaying additional data on semantic links beyond the current semester. For example, the map shown in Fig. 5c additionally displays the data on the links between the learning topics of the third semester and the other levels of the hierarchy shown in Fig. 1b (nodes D1, D2, D4 and Dip). Here those links are additionally highlighted in red, which allow to strengthen the argumentation for learning topic t12 in relation to future learning elements (from t6 and t12 to D4). Now let us turn to the data of the same student in the functional aspect on anthropomorphic UGVA images. We will superimpose the contribution values of grades for those disciplines, which correspond to each functional part of the image (S) on the invariant image of the Informatics and Computer Science program shown in Fig. 4a. As a scope we chose values from red (grades tend to zero) to green (grades tend to 100), with an

Cross-Cutting Visual Support of Decision Making

11

uncertainty zone of 50 (white). For example, the forearm of the right hand (zone S3) corresponds to the skill group “To master mathematical and algorithmic methods of information processing” which is poorly mastered (weighted score of 54.2 points corresponds to almost white color from Fig. 4b). In the competency context (Fig. 4c) the mastery of the competencies is scored at 37.3 points (if the confidence coefficient is given to points in the interval 0–100), which corresponds to a dark pink color (see [12] for details). When analyzing the problems of the learning situation from the functional point of view, it is necessary to point out the insufficient development of competencies from the S3 zone (core part) and explain the reasons to the student (including the argumentation from the structural aspect presented on the CMKD above). By combining maps and images for different aspects an atlas is formed, reflecting the current state of the learning situation, on the basis of which verification of a wide range of hypotheses can be accelerated in the process of ITS operation, and the corresponding decisions can be explained. A limitation of the proposed visualization scheme for ITS is the insufficient expressiveness of CMKDs and UGVA images for the nano level and meta level scopes (see Fig. 1a). The nano level involves synthesizing the dialogue in a subject logic rather than in a methodological logic. In this situation, an ontological approach (for example, as proposed in [9]) will be useful. On the other hand, using the meta-level is not well elaborated in the publications on ITS and requires additional research. In particular, student data in the transition between educational levels goes beyond the scope of individual tutoring systems and involves the development and standardization of mechanisms for the migration of the digital educational footprint between ITSs.

4 Conclusion Making methodological decisions by Intellectual Tutoring System (ITS) mechanisms needs verifiable approaches to both selecting the control activity [5] and explaining it to a person. Without trust between the tutoring system and the student, a complete Personalized Learning Environment (PLE) cannot be formed [10]. Therefore, research on decision-making by ITS mechanisms and decision explanations will undoubtedly arouse increased interest in the community. The proposed cross-cutting approach based on a combination of graphic notations such as CMKD and UGVA, shows very interesting results. They have yet to be thoroughly analyzed in order to show their impact on the e-learning process as a whole. Current areas of our work on the use of cross-cutting visual support in the learning process are: – monitoring the learning situation of master’s degree students during the whole period of study (from the entrance exams to the thesis defense); – recording students’ reactions to the system’s visual support of recommendations and their explanations as well as evaluating their impact on the academic performance changes; – expanding the variety of training programs to apply the described approach and collecting sufficient experimental data.

12

V. Uglev and T. Gavrilova

The data obtained in this study suggest that the cross-cutting support of the decisionmaking process and explanation of decisions can lead to a significant expansion of ITS capabilities. This should help to increase the level of confidence in ITS decisions in the context of the XAI concept.

References 1. Anokhin, P.K.: The functional system as a unit of organism integrative activity. In: Mesarovi´c, M.D. (eds.) Systems Theory and Biology, pp. 376–403. Springer, Heidelberg (1968). https:// doi.org/10.1007/978-3-642-88343-9_15 2. Arrieta, A.B., et al.: Explainable Artificial Intelligence (XAI): concepts, taxonomies, opportunities and challenges toward responsible AI. Inf. Fusion 58, 82–115 (2020) 3. Baranova, S., Stepanova, E., Kalinovsky, K., Sokolov, I., Khrabrova, N.: Automated Analysis and Synthesis of University Curricula Based on an Array of Didactic Units, vol. 3, pp. 216, 221. Bulletin of Krasnoyarsk State Agrarian University (2014) 4. Filimonov, V.: Method for cognitive visualization multi-parameter system components. In: XIII International Conference on Robotics and Artificial Intelligence, pp. 161–172 (2021). (in Russian) 5. Krouska A., Troussas, C., Giannakas, F., Sgouropoulou, C., Voyiatzis, I.: Enhancing the effectiveness of intelligent tutoring systems using adaptation and cognitive diagnosis modeling. In: Novelties in Intelligent Digital Systems: Proceedings of the 1st International Conference (NIDS 2021), Athens, Greece, 30 September – 1 October 2021, vol. 338, p. 40. IOS Press (2021). https://doi.org/10.3233/FAIA210073 6. Grann, J., Bushway, D.: Competency map: visualizing student learning to promote student success. In: Proceedings of the Fourth International Conference on Learning Analytics and Knowledge, pp. 168–172 (2014) 7. Kossiakoff, A., Sweet, W., Seymour, S., Biemer, S.: Systems Engineering Principles and Practice. Wiley-Interscience, New York (2011) 8. Sychev, O., Denisov, M., Anikin, A.: Verifying algorithm traces and fault reason determining using ontology reasoning. In: CEUR Workshop Proceedings, pp. 49–53 (2020) 9. Takada, S., et al.: Toward the visual understanding of computing curricula. Educ. Inf. Technol. 25(5), 4231–4270 (2020). https://doi.org/10.1007/s10639-020-10127-1 10. Uglev, V.: Evaluate curricula balance for software engineering education with using UGVA method. Mod. Inf. Technol. IT-Educ. 17(3), 684–696 (2021). https://doi.org/10.25559/SIT ITO.17.202103.684-696. (in Russian) 11. Uglev, V., Sychev, O.: Concentrating competency profile data into cognitive map of knowledge diagnosis. In: Basu, A., Stapleton, G., Linker, S., Legg, C., Manalo, E., Viana, P. (eds.) Diagrams 2021. LNCS (LNAI), vol. 12909, pp. 443–446. Springer, Cham (2021). https://doi. org/10.1007/978-3-030-86062-2_46 12. Uglev, V., Sychev, O., Gavrilova, T.: cross-cutting support of making and explaining decisions in intelligent tutoring systems using cognitive maps of knowledge diagnosis. In: ITS 2022. LNCS, vol. 13284, pp. 51–64 (2022). Springer, Cham. https://doi.org/10.1007/978-3-03109680-8_5 13. Zouri, M., Ferworn, A.: An ontology-based approach for curriculum mapping in higher education. In: 2021 IEEE 11th Annual Computing and Communication Workshop and Conference (CCWC), pp. 0141–0147. IEEE (2021) 14. Zykina, A., Kaneva, O., Munko, V.: The development of approaches for obtaining automated solution on the formation of the curriculum. Mod. Inf. Technol. IT-Educ. 14(4), 931–937 (2018). https://doi.org/10.25559/SITITO.14.201804.931-937

Personalized Learning in an Intelligent Educational System Valentina Terzieva1

, Tatyana Ivanova2(B) , and Katia Todorova1

1 Institute of Information and Communication Technologies, Bulgarian Academy of Sciences,

Bl. 2, Acad. G. Bonchev Street, 1113 Sofia, Bulgaria {valentina.terzieva,katia.todorova}@iict.bas.bg 2 Technical University of Sofia, Sofia, Bulgaria [email protected]

Abstract. The growing expansion of smart devices and intelligent technologies encourages the development of intelligent learning environments where a personalized and adaptive learning process is offered. In this regard, we perform research on how to implement appropriate tutoring strategies to provide such a learning process in an Intelligent Educational System (IES) context. The paper discusses good practices in the conceptual model development of an IES within the project ISOSeM. Our research proposes a Knowledge Model for IES that can support dynamic learning path-based personalization. This model includes ontologically represented information about prerequisites, learners, learning goals, teaching strategies, course, educational content resources, and assessing learners’ knowledge. The approaches to an automatic selection and generation of personalized learning paths using the proposed model are considered. The goal is to accommodate tutoring to students’ real-time learning advance and make learning paths, activities, and resources meet learners’ individual needs and preferences. Keywords: Intelligent educational system · Personalized learning · Personalized learning path · Knowledge model · Ontology

1 Introduction The current information age has changed the way education has occurred. Accordingly, the teachers’ and students’ needs, preferences, and expectations also have changed considerably. They require the use of more innovative, supportive educational technologies. The technologies can help meet these necessities and realize the new information-age paradigm of education – smart or intelligent education [1, 2]. It demands a new approach to students’ assessment considering what knowledge the students have already acquired and what they have to acquire according to the learning objectives. Personalized learning is another feature of smart education. Such an approach to learning requires building models of students by gathering data about their characteristics, goals, preferences, achievements, etc. [3, 4]. In the framework of smart education, where personalized learning is performed, students have individual learning goals to meet. Since © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 13–23, 2023. https://doi.org/10.1007/978-3-031-17601-2_2

14

V. Terzieva et al.

that students’ learning styles, goals, interests, and paces of learning vary significantly, the intelligent educational systems (IES) have to assist teachers in the adaptation of learning paths and learning content as per the students’ needs [5]. Within IES, a vast amount of data is gathered through smart devices [6, 7], and their proper and real-time analysis is of crucial importance. Efficient data processing supports prompt decision-making for the quality improvement of teaching. Among the used knowledge-based data processing technologies are data mining (DM), artificial intelligence (AI), big data techniques, machine learning, pattern recognition, learning analytics (LA), artificial intelligence (AI), semantics, ontologies, neural networks (NN), etc. [8–10]. Following these tendencies, our team has initiated the research project ISOSeM [11], which aims to model the architecture and functions of an intelligent educational system (IES) that provides personalized and adaptive educational services. This paper presents an approach to practical personalization of the learning process in ISOSeM IES. It proposes personalized teaching based on data gathered by the educational system and sophisticated algorithms for defining and implementing the best tutoring strategy. A knowledge model representing information about all teaching and learning-related issues and supporting dynamic learning path-based personalization in IES is developed.

2 Related Works The idea of personalized learning has appeared for many decades. Personalized learning can be considered as an umbrella term for different customized educational approaches that vary significantly depending on the context, objectives, applied pedagogy, used technology, and other factors [12]. The recent growth of innovative technologies has created conditions that facilitate personalized education [13]. Many advantages of AI techniques employed to achieve more intelligent adaptive learning environments are discussed in [14]. Usually, user-adaptive educational systems provide personalized educational experiences based on user modeling [3]. IES, innovative virtual learning, and data analysis and prediction are the main applications of the AI approach in education. AI-based techniques for learning analysis, recommendation, knowledge understanding, and acquirement, based on machine learning, DM, and knowledge modeling, are used in educational systems [15]. Student models or profiles include essential information gathered by direct input from the user and tracing learning interactions, identifying learning styles and preferences, etc. An approach for recommending learning paths using AI techniques (Ant colony optimization) for children facing learning disabilities is presented in [16]. A learning path recommendation system based on a variable-length genetic algorithm by considering learners’ learning styles and knowledge levels is shown in [17]. The effectiveness of this algorithm is tested in a practical e-learning environment. Personalized learning can be achieved through adaptation in diverse modes – personalizing the interface, enabling selection of individual learning paths, supplying different learning content, varied presentation styles, educational games [18], etc. Approaches for adapting learning content to students’ characteristics (learning style, task performance, cognitive load, etc.) are proposed in [19]. Intelligent tutoring systems use the same parameters for selecting learning paths and tasks. Automated course sequencing is essential in Intelligent Tutoring Systems (ITS) because it can determine the personalized learning path of every student. There are many

Personalized Learning in an Intelligent Educational System

15

attempts to generate or automatically select learning paths in the literature. Research in this area uses semantic models of one or more parts of the learning process, such as the learner model, curriculum and learning content model, context model, etc. E-assessment is the primary source of information about students’ knowledge and skills. Research [20] presents AONet, an ontology network that conceptualizes the eassessment domain. It is used for supporting the semi-automatic generation of assessment. An ontology network is not exactly a set of interconnected individual ontologies. In an ontology network, The DOOR (descriptive ontology of ontology relations) ontology is used for describing meta-relationships (e.g., includedIn, equivalentTo, similarTo) among the networked ontologies. AONet consists of five ontologies: course topic domain, educational resources, and assessment. An ontology-driven approach that uses a learning path adaptivity mechanism is presented in [21]. Two ontologies and rules are applied – domain ontology, modelling mathematical knowledge, a learning path-adaptation task ontology, and a set of semantic rules, which enable the learning path adaptivity. It is shown that an ontology-driven approach aiming to specify a learning path for learners to return to needed prerequisite learning can facilitate e-learning. A framework for inferring adapted and contextual learning paths is proposed in [22]. It is based on contextual graph approaches (Ontology and Context Dimension Tree) and Bayesian Networks. This framework can select contents and services useful during visits in real scenarios such as archaeological parks or museums according to the learner’s profile and the context. The system can design and suggest personalized learning paths for improving training using ontological models and predictive techniques. Research [23] presents a conceptual model of an intelligent system based on semantic technologies and learning paths. This system supports self-regulated learning. The system measures student progress on different learning levels for learning path generation. External information from the internet also can be used to support the generation of learning paths. An approach for determining the sequence of learning concepts during the course using ontology and Wikipedia information is proposed in [24]. A textmining algorithm is applied to extract knowledge from Wikipedia to support determining the sequence of learning objects. Concepts and relationships stored in the ontological model of course content and knowledge about the same or related concepts extracted from Wikipedia articles are used for determining the best sequence for introducing new concepts and discussing interconnect relations. Another approach for proposing personalized recommendations of learning paths to students is based on Big Data [25]. The paper presents research aimed at improving the quality of hybrid and online learning. This research has shown an online learning path model by exploring the big data of online learning processes. The model is based on data about students’ learning habits, and its primary goal is to find excellent learning paths. Frequently, these learning paths have been classified as paths for five learning styles: discovery learning, discursive learning, exploratory learning, cooperative learning, and task-based learning. This brief survey assumes that the creation and usage of learning paths are significant for organizing almost all types of personalized learning, from web-based to IMS- based, from formal to informal, and from self-regulated to teacher-driven. Another conclusion is that diverse information (about learners, learning styles, learning content, assessments, and learning strategies) is needed to achieve automated generation and selection of

16

V. Terzieva et al.

personalized learning paths. The semantics-based organization of these metadata can ensure the best flexibility and possibilities for automation of personalized learning path generation and selection in intelligent tutoring systems.

3 Intelligent Educational System in the Context of ISOSeM Considering requirements from the related research, the authors have implemented many good practices in the development of the conceptual model of an IES [26] within the project ISOSeM. This IES utilizes smart devices such as the Internet of Things (IoT) – sensors, actuators, wearable devices, etc. It also integrates intelligent technologies, including DM, LA, AI, NN, semantics, and ontologies. These technologies enable IES to track and personalize physical environment parameters and learning process flow. The physical parameters in the classroom environment (temperature, lighting, humidity, oxygen level, etc.) that can affect the learning process should be monitored and controlled [26]. Tracking the learning process flow includes recognizing students’ behavior and interactions within the learning system. All the data are collected and processed accordingly by the pedagogical module. This module applies educational data mining, big data techniques, LA, and AI [14] to select and provide appropriate tutoring strategies and learning resources to ensure a personalized educational process. All analytics results and derived information are the basis for building and updating learners’ profiles. They include data about characteristics of students – personal, administrative, educational (knowledge level, learning style, learning goals, academic achievements, etc.), and cognitive (perception, memory, concentration, etc.). These data support the decision-making process to determine the most appropriate teaching approaches, strategies, learning resources, learning paths, etc., to provide efficient education. Tutors refer to these profiles when dynamically recommending suitable learning paths and content according to the student’s knowledge and progress, following the students’ learning goals. Thus, a personalized learning experience that reflects students’ preferences and interests is offered. As a result, an intelligent learning process comprising a personalized and adaptive learning curriculum, learning paths, and learning resources is provided. Students are offered continuous feedback that brings a more flexible, engaged, and motivating learning experience. The learning process becomes more efficient and individualized [1, 13]. Some research concerning personalized tutoring uses the ontology-based description of tutoring content or learning objects. Others pay attention mainly to learner modeling or approaches to use intelligent technologies to control the tutoring process. But we could not find a comprehensive knowledge model for a structured representation of all the knowledge needed for conducting a personalized tutoring process. The following section explains our view about modeling knowledge necessary for providing dynamic learning path-based personalization. It presents a general knowledge model and approach to achieving a personalized learning experience by developing and using learning paths within ISOSeM IES.

Personalized Learning in an Intelligent Educational System

17

4 Personalization in ISOSeM IES 4.1 The Knowledge Model for Dynamic Learning Path-Based Personalization Personalization of the learning process in a particular course can be done by proposing learning content according to the needs of every learner or by offering a specific format of the content presentation to simplify learning or increase motivation. Presentation-based personalization usually uses learner models and pedagogy, saying what type of content presentation is best for particular characteristics of learners. A prerequisite is anything you need to know, understand, or do before learning something new. Every learning content has some prerequisites, although it is not easy to explicitly specify them all. Content-based personalization is closely related to the prerequisites. Thus, to create or find the best personalized content-based learning path, it is necessary to include content related to these prerequisites and to the learner’s needs. One of the main approaches to the personalization of learning is the generation and usage of personal learning paths. It is a learner-centered approach based on information about the learner (specific knowledge, goals, individual and psychological characteristics, or preferences) to propose appropriate learning content. A personal learning path is a sequence of learning content elements, exercises, or activities that a particular learner should do to achieve some learning goal. In personalized learning, learners can choose their path, i.e., learning content, activities, and assignments. Otherwise, an e-learning system can recommend an appropriate learning path according to the learner’s characteristics. Learning paths can be generated statically by teachers or dynamically by the system. A dynamic and flexible generation and a further selection of learning paths are necessary to achieve real personalization. Also, there is a need for rich knowledge models of learners and learning resources metadata. As discussed, much research explores the usage of ontologies for personalization, resource searching, and recommendation in elearning systems. We could not find a complete knowledge model supporting these tasks yet, so we propose an ontology-based knowledge model suitable for content-based learning path generation or selection. This model includes ontologically-represented information about course content, prerequisites, learners, learning goals, teaching strategies, educational resources, and assessing learners’ knowledge. Course content ontology is a semantic model of learning content in a particular course. It classifies concepts defined in this course and offers its model according to the definitions and relations presented. In such an ontology, learning objects’ names are added as instances to make recommendations of learning content for learning a specific knowledge. Prerequisites ontologies represent a semantic model of learning content that learners need to know before a new course. They classify concepts that can be the subject of several previous courses. In these ontologies, additional resources’ names are added as instances to make a viable recommendation of content to learn knowledge of previous courses. We recommend using a system of mapped ontologies for modeling prerequisites because prerequisites can come from several other courses, and it can be possible to reuse its semantic models. Learners’ ontologies are learner profile ontologies, where characteristics and actual knowledge of learners are stored. These ontologies are semantic representation of learner models. We recommend using a system of mapped ontologies for modeling user profiles.

18

V. Terzieva et al.

The reasons are that adding new learners to the course can lead to the addition of different types of learners whose properties have already been modeled in other existing ontologies. Leaning goals ontology classifies course tutoring goals. Learning goals are defined in courses or attached to the learning objects to map the learner’s goals, tutoring goals, and learning objects. Teaching strategies ontology ensures the reusability of teaching strategies by proposing its semantic classification and description. Reusable tutoring strategies ensure the maximal separation of pedagogical behavior from domain knowledge. Every intelligent tutoring system needs to implement all the strategies usable for teaching concept understanding, problem-solving, and all other sub-processes of the tutoring process, as diverse strategies can be the most useful for different students. Tutoring strategies should be systematized properly according to many dimensions: domain-dependence, globalism, relation to pedagogical theories, principles, activities, personalization level, types of used learning objects, learning goals, etc. Strategies should be distinctly separated as Domain–independent and domain-dependent to support their reuse. Teaching strategies ontology and learning goals ontology should be mapped to denote which teaching strategies are appropriate for achieving a specific goal. The Educational resources ontology describes the educational resources used in the e-learning system. There are several standards in e-learning to describe educational resources (including Dublin Core, IMS LOM, SCORM, IMS Learning Resource Metadata), and most of them are constantly evolving. Many science projects also have presented ontologies, conceptualizing most of these standards. Some of these ontologies can be reused in intelligent educational systems. Evaluation of the knowledge and skills of the learners is very important for dynamic adaptation of tutoring, generation, and usage of adequate learning paths [27, 28]. An ontological representation of the assessment domain is essential for the flexible and dynamic organization of the assessment. The Assessment ontology is a semantic model of description of all the instruments used to evaluate learning processes and learners’ knowledge, skills, motivation, etc. This ontology describes the assessment process, tasks, activities, and instruments. For supporting the efficient usage of these ontologies in the educational system, semantic alignments between ontologies are needed (Fig. 1).

Fig. 1. Knowledge model to support dynamic learning path-based personalization.

Personalized Learning in an Intelligent Educational System

19

The proposed knowledge model is used by the software components of the IES (services or intelligent agents, implementing algorithms for searching, recommendation, selection, sequencing, or management of learning resources). Semantic annotation of learning objects based on the proposed knowledge model is significant for proper selection, sequencing, or recommendation. IES should store and use a set of rules to support intelligent tutoring. For example, in what situations, defined by the knowledge stored in the ontological model, to recommend external or internal resources; what learning path the learner should follow, when the learner can pass to the next module, or when to repeat some activities, etc. 4.2 Development and Usage of Learning Paths A learning path comprises a sequence of learning content resources that all the students in a course should learn to complete some learning goal. The personal learning path is intended for a particular learner to achieve a personal learning goal. Teachers can manually develop learning paths and store them as resources for future usage. A variety of learning paths is needed to achieve the learning goals in personalized learning. Both the development and selection of the appropriate learning path manually are complex. We will discuss how knowledge models, like the one proposed above, can be used by IES to generate automatically and select suitable personalized learning paths. Generally, the following steps need to be performed to develop a learning path: 1. Identify the tutoring goals, outcomes, or objectives. 2. Identify resources, tasks, and activities to be included in the learning process. 3. Specify the sequence of selected resources, tasks, and activities. Initially, teachers should specify and store some set of general learning paths considering the goals of the course, tutoring strategies, and available resources. These learning paths have to be annotated using teaching strategies ontology, educational resource ontology, learning goals ontology, and course content ontology to ensure automated usage by the IES. When the course begins, a General tutoring strategy that is responsible for all tutoring activities is applied to check the students’ prior knowledge. Based on that, the necessary external resources are searched and described for usage by particular learners. The General tutoring strategy for tutoring during one module is: 1. Every learner does a test about prerequisites; 2. Test results are analyzed, and problems related to the past knowledge and skills of every learner are detected; 3. Stored learning paths are checked, and if there is an appropriate learning path, it is used during tutoring the specific learner; 4. If an appropriate learning path is not found, a new learning path is generated, stored, and used for tutoring. Then personal learning paths for every learner are generated automatically by the IES by modifying general learning paths using stored rules. The specification of the needed external content is added in the correct places to ensure the gaps filling in prerequisite

20

V. Terzieva et al.

knowledge for every learner. After completing the course module, every learner does a test on this module. A specific teaching strategy for additional tutoring is selected based on the test results. Rule-based techniques and the proposed ontological model are used to find what causes the learner’s knowledge gaps. The main questions are: At what level are the learner’s problems (understanding, memorization, reasoning, doing something, etc.)? If there are problems, what are they related to – the prerequisites or the course knowledge? What indeed causes these problems (low motivation, inadequate skills, etc.)? A personal learning path also should blend the e-learning course essentials with the needs and preferences of each learner. For the creation of personal learning paths in IES, the following issues are significant: • General learning goals; • Individual learner’s goals and objectives; • Assessment and self-assessments – used as dynamic sources of information about learner’s knowledge and skills; • Periodical milestones to give learners the chance to check their progress; • Various presentation formats according to different learning styles of learners; • Immediate and constructive feedback. Personal learning paths are annotated by teaching strategies ontology, educational resource ontology, learning goals ontology, course content ontology, and learner ontology to ensure their automated usage by the IES and are stored for future usage. Rules are used both during the generation and selection of appropriate personal learning paths. We use the Semantic Web Rules Language (SWRL) to generate and represent rules. A query language is needed to extract information from OWL ontologies.

5 Discussion and Conclusion The recent pandemic pushes the enhancement of distance learning and has changed the way students acquire knowledge. Intelligent educational systems have gained more popularity. The presented IES aims to provide smart educational services to digital generation students thanks to the implementation of intelligent technologies. During tutoring, there are many situations when detailed knowledge about learners, learning content, and pedagogy is essential for IES to conduct successful learning for every student. For example, when a student has low results on a test, he should repeat some learning activities. ITS should recommend which learning resources to be used in what order (e.g., to change the learning path). A good recommendation can be made only based on the knowledge about relations between learning resources. The learner’s psychological properties, preferences, or capabilities for learning are also important. Some of the needed learning content for a particular learner cannot be part of the tutoring course (e.g., when a learner has some learning disabilities or did not meet some course prerequisites). In this case, knowledge about interoperability between ITS is necessary to support automatic finding and recommendations of external resources. The usage of many small ontologies encourages knowledge reuse and automatical evolution

Personalized Learning in an Intelligent Educational System

21

of metadata, in correspondence, changes in the elements of the e-learning system. Some ontologies describing course content, for example, can be easily embedded in the ontological system of another related course as a description of external resources. Local changes to the course content model can be made whenever learning difficulties arise to support understanding of the learning content, e.g., by visualization of relationships between concepts or their properties or by supporting the recommendation of suitable learning content. Possibilities for reuse and modularity decrease development time and labor for this complex semantic model. At the same time, this model proposes sufficient knowledge needed for the dynamic generation and selection of elaborated personalized learning paths. The goal is to accommodate tutoring to students’ real-time learning advance and make the learning paths, activities, and resources meet students’ needs. To sum up, this paper introduces a new knowledge model for personalized and adaptive learning within the framework of IES. The presented semantic model is based on ontology networks and alignment of ontologies, used for representing complex knowledge about learners, pedagogy, and learning content. The proposed ontology-based knowledge model can systematically store all this knowledge, support its updates, and make it usable for software services or intelligent agents responsible for conducting intelligent tutoring and learning. Also, the paper presents and discusses a methodology for creating personalized learning paths and storing them in the repository for multiple usages. The stored rules support selecting the best learning path from the repository or making dynamic changes to the previously selected learning paths considering the knowledge about the learner’s success or satisfaction. Acknowledgements. The authors would like to thank the Research and Development Sector at the Technical University of Sofia for the financial support. This research is supported by the Bulgarian FNI fund through the project “Modeling and Research of Intelligent Educational Systems and Sensor Networks (ISOSeM)”, contract KP-06-H47/4 from 26.11.2020.

References 1. Peng, H., Ma, S., Spector, J.M.: Personalized adaptive learning: an emerging pedagogical approach enabled by a smart learning environment. Smart Learn. Environ. 6(1), 1–14 (2019). https://doi.org/10.1186/s40561-019-0089-y 2. Zhu, Z.-T., Yu, M.-H., Riezebos, P.: A research framework of smart education. Smart Learn. Environ. 3(1), 1–17 (2016). https://doi.org/10.1186/s40561-016-0026-2 3. Brusilovsky, P., Millán, E.: User models for adaptive hypermedia and adaptive educational systems. In: Brusilovsky, P., Kobsa, A., Nejdl, W. (eds.) The adaptive web. LNCS, vol. 4321, pp. 3–53. Springer, Heidelberg (2007). https://doi.org/10.1007/978-3-540-72079-9_1 4. Okoye, K.: A systematic review of process modelling methods and its application for personalised adaptive learning systems. J. Int. Techn. Inf. Manag. 27(3), 23–46 (2019) 5. MacLeod, J., Yang, H., Zhu, S., Li, Y.: Understanding students’ preferences toward the smart classroom learning environment: development and validation of an instrument. Comput. Educ. 122(7), 80–91 (2018) 6. Ilchev, S., Andreev, R., Ilcheva, Z.: HybridNET management and sensor data acquisition system. In: Proceedings of 7th International Conference on the Internet of Things, pp. 1–2 (2017)

22

V. Terzieva et al.

7. Ivanova, M., Boneva, A., Ilchev, S.: Learning performance facilitation in a sensor-based intelligent classroom. In: Proceedings of BdKCSE, pp. 1–8. IEEE (2021) 8. Gros, B.: The design of smart educational environments. Smart Learn. Environ. 3(1), 1–11 (2016). https://doi.org/10.1186/s40561-016-0039-x 9. Rico-Bautista, D., Medina-Cardenas, Y., Coronel-Rojas, L.A., Cuesta-Quintero, F., MaestreGongora, G., Guerrero, C.D.: Smart university: key factors for an artificial intelligence adoption model. In: García, M.V., Fernández-Peña, F., Gordón-Gallegos, C. (eds.) Advances and Applications in Computer Science, Electronics and Industrial Engineering. AISC, vol. 1307, pp. 153–166. Springer, Singapore (2021). https://doi.org/10.1007/978-981-33-4565-2_10 10. Radosavljevic, V., Radosavljevic, S., Jelic, G.: Ambient intelligence-based smart classroom model. Interact. Learn. Environ. 30(2), 307–321 (2022) 11. ISOSeM Homepage. http://isosem.bas.bg/ 12. Groff, J.S.: Personalized Learning: The state of the field & future directions, Center for Curriculum Redesign (2017) 13. FitzGerald, E., Jones, A., Kucirkova, N., Scanlon, E.: A literature synthesis of personalized technology-enhanced learning: What works and why. Res. Learn. Technol. 26 (2018). http:// dx.doi.org/10.25304/rlt.v26.2095. Article No. 2095 14. Colchester, K., Hagras, H., Alghazzawi, D., Aldabbagh, G.: A survey of artificial intelligence techniques employed for adaptive educational systems within e-learning platforms. J. Artif. Intell. Soft Comput. Res. 7(1), 47–64 (2016) 15. Chen, L., Chen, P., Lin, Z.: Artificial intelligence in education: a review. IEEE Access 8, 75264–75278 (2020) 16. Roman, J., Mehta, D.: Intelligent content sequencing through preferences and feedback of special need learner (2019) 17. Dwivedi, P., Kant, V., Bharadwaj, K.K.: Learning path recommendation based on modified variable length genetic algorithm. Educ. Inf. Technol. 23(2), 819–836 (2017). https://doi.org/ 10.1007/s10639-017-9637-7 18. Antonova, A., Dankov, Y., Bontchev, B.: Smart services for managing the design of personalized educational video games. In: BCI’2019 Proceedings, pp. 1–8, ACM (2019). Art. No 20 19. Van Seters, J., Ossevoort, M., Tramper, J., Goedhart, M.: The influence of student characteristics on the use of adaptive e-learning material. Compt. Educ. 58(3), 942–952 (2012) 20. Romero, L., North, M., Gutiérrez, M., Caliusco, M.: Pedagogically-driven ontology network for conceptualizing the e-learning assessment domain. In: International Forum of Educational Technology and Society (2015) 21. Chi, Y.-L., Chen, T.-Y., Tsai, W.-T.: Creating individualized learning paths for self-regulated online learners: An ontology-driven approach. In: Rau, P.L.P. (ed.) CCD 2014. LNCS, vol. 8528, pp. 546–555. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-07308-8_52 22. Colace, F., De Santo, M., Lombardi, M., Mosca, R., Santaniello, D.: A multilayer approach for recommending contextual learning paths. J. Internet Serv. Inf. Secur. 10(2), 91–102 (2020) 23. Romero, L., Saucedo, C., Caliusco, M., Gutiérrez, M.: Supporting self-regulated learning and personalization using ePortfolios: a semantic approach based on learning paths. Int. J. Educ. Technol. High. Educ. 16(1), 1–16 (2019). https://doi.org/10.1186/s41239-019-0146-1 24. Ramírez-Noriega, A., Juárez-Ramírez, R., Jiménez, S., Martínez-Ramírez, Y., Figueroa Pérez, J.: Determination of the course sequencing to intelligent tutoring systems using an ontology and Wikipedia. J. Intell. Fuzzy Syst. 34(5), 3177–3185 (2018) 25. Liu, Z., Dong, L., Wu, C.: Research on personalized recommendations for students’ learning paths based on big data. Int. J. Emerg. Technol. Learn. (iJET) 15(8), 40–56 (2020) 26. Terzieva, V., Todorova, K., Ivanova, T.: Conceptual model of intelligent educational system and the need of big data analytics. In: Proceedings of BdKCSE’2021. IEEE (2021)

Personalized Learning in an Intelligent Educational System

23

27. Ouahabi, A., Lequeu, T.: Remote knowledge acquisition and assessment during the COVID-19 Pandemic. Int. J. Eng. Pedagog 10(6), 120–138 (2020) 28. Kanetaki, Z., Stergiou, C., Bekas, G., Troussas, C., Sgouropoulou, C.: Evaluating remote task assignment of an online engineering module through data mining in a virtual communication platform environment. Electronics 11(1), 158 (2022)

Electronic-Service Learning to Sustain Instruction with Civic Engagement During the COVID-19 Pandemic Aurelio Vilbar(B) University of the Philippines Cebu, Lahug, Cebu City, Philippines [email protected]

Abstract. The COVID-19 pandemic disrupted higher education courses integrating face-to-face (FtF) community-based service learning into their curriculum. There is a need to transform FtF service-learning (FtF-SL) into electronic service-learning (e-SL) to promote safety. e-SL studies show positive results in academic enhancement but reveal struggles sustaining online collaboration. This paper reports the experience of graduate students who took the e-SL as an alternative assessment in production of in structural materials course during the pandemic. The students responded to the public school’s request in Olango Island, Cebu, Philippines, to virtually coach their teachers in designing their remedial instructional materials. Findings revealed that the project promoted student academic enhancement, civic engagement, personal growth and developed the public-school teachers’ research skills. The online conferencing platforms and online messenger sustained the collaboration despite no FtF meetings. The Experts’ Evaluation showed that the materials passed all criteria, and pilot testing results showed that the users described the materials as interesting, educational, and enjoyable. Keywords: Electronic-service learning · Online collaboration · COVID-19 pandemic

1 Rationale COVID-2019 pandemic disrupted the delivery of instruction and authentic assessment which forced academic institutions to shift to remote teaching and learning [1, 2]. This unprecedented shift posed challenges, especially to graduate education courses integrating face-to-face (FtF) service-learning as their teaching approach before the pandemic [3]. FtF service-learning is a course-related service activity conducted by the students that meet the community’s needs [4]. Anchored on transformative education, it facilitates knowledge and holistic care through organized community work [5]. The immersion component of service learning allows the students to apply the course objectives in their partner community [6]. Because it uses the frameworks of Dewey’s learning-by-doing principle and Freire’s humanization concept, service-learning is a sound humanistic approach to teaching [4]. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 24–32, 2023. https://doi.org/10.1007/978-3-031-17601-2_3

e-SL to Sustain Instruction with Civic Engagement

25

However, due to the FtF component of service-learning, it could not be during the pandemic, for it could jeopardize the safety of the students and the community. There is a need to transform FtF-SL into electronic-service learning (e-SL). e-SL is technologymediated, whose instructional and service components are delivered online [7] or offline [8]. However, e-SL has limited studies to gauge its effectiveness [9]. Like FtF-SL, e-SL can develop academic learning, personal growth, professional growth, and community engagement [10]. However, reports showed students had initial struggles in collaborating with the community due to lockdowns and the less engaging and authentic virtual format [11]. This research examined the usefulness of e-SL in teaching the graduate course “Production/Adaptation and Evaluation of Language Learning Materials” without FtF interaction during the pandemic. In this project, the students addressed the request of a public school in Olango Island, Cebu, to virtually coach its teachers in developing remedial programs with instructional materials. This school community has been the university partner before the pandemic. Sustaining the volunteerism project during the health crisis allowed the university to attain its public service mission [12, 13]. This research aimed to determine the (1) impact of using e-SL on the graduate students’ course content, personal growth, and civic engagement; (2) the impact on the teachers’ skills in conducting action research by creating remedial programs; and (3) the technology used by the participants to sustain collaboration during the pandemic.

2 Literature Review e-SL is an alternative pandemic teaching to achieve the course objectives and community engagement without violating the health protocols [14]. As a transformative education, e-SL is technology-mediated, whose instructional and service components are delivered online or offline [7], depending on the context of the learning outcomes and the community. Despite the absence of the FtF component, e-SL has retained a positive impact of FtF-SL in promoting academic learning, personal growth, professional growth, and community engagement [9, 10, 15]. It can promote student engagement and morale within their online learning classroom community [16]. e-SL promoted course content and enjoyment, as reported by the students who had assignments in developmental psychology to create electronic storybook recordings and educational videos for a local library [9]. Students claimed that e-collaboration promoted flexibility and more control over the time and place of their assignments. Aside from the community partnership and curriculum design [17], what was crucial in the e-SL during the pandemic was the use of effective and engaging online-offline technology platforms to sustain meaningful collaboration between the students and the community [18–20]. In the Project WeCan, [18], students volunteered to help government-aided secondary students with a general academic performance by creating online educational videos related to their courses. Findings proved that e-SL developed their life appreciation, satisfaction, empathy, and knowledge. The subsequent studies provided these suggestions on how to promote a successful e-SL implementation: create a hybrid pedagogy and instructional design that reflect principles of learning, good practice, committed partnership with the teachers, students, and communities [21], design a collaborative

26

A. Vilbar

environment sensitive to the community’s needs [22] but remaining flexible to the students’ context; and use engaging technology and software communications for student volunteers and teachers [7].

3 Methods 3.1 The e-Service Learning Project The course was taught in a fully online remote learning from September 2020 to November 2020 at a state university in Cebu. Olango Island was the partner community where the e-SL was conducted. The school requested e-coaching due to the COVID-19 cases in Cebu province [23]. Twelve teachers volunteered to participate in the project. Their need to create remedial programs matched with the course objectives. All graduate students volunteered to coach the teachers in creating their instructional materials. The online coaching was conducted in small groups. The four groups were the following: Game-Based Reading Remedial Program, Blended Reading Programs, Storybooks, and Teacher-Made Math Tutorial Videos Using Translanguaging. The University of the Philippines Cebu Ugnayan ng Pahinugod (the Volunteer Service Arm of UP) co-organized the project. 3.2 Data Collection and Analysis This study used the exploratory design [24], in which various data gathering procedures were utilized to address the phenomenon of conducting e-SL during the pandemic. It used online anonymous open-ended surveys, reflections, semi-structured interviews, and focus group discussions (FGD) to determine the impact of e-SL on the graduate students and the teachers. This research was anchored on the Critical Service-Learning Approach [25], which situates service-learning as a tool for social change orientation, redistribution of power, reciprocity, and developing authentic relationships between the student service learners and the partner community. All participants submitted reflections. Reflections as mental activities can allow the students to critically examine their experience [26] in the e-SL project. The open-ended survey and reflections were analyzed using Harding’s coding thematic analysis framework [27]. Another professional from UP Cebu Ugnayan ng Pahinungod did the coding using the framework in categorizing academic enhancement, personal growth, and civic engagement [28].

4 Results and Discussion 4.1 Project Promoted Students’ Academic Enhancement Table 1 shows that e-SL developed the students’ academic enhancement, personal growth, and civic engagement, as shown in their reflections. The students claimed they gained knowledge and skills in materials production and applied theories in a real context. Student A answered, “I created reading materials that can help public school students.” Student B added, “We helped teachers make a workbook for the Grade 7 students who

e-SL to Sustain Instruction with Civic Engagement

27

Table 1. Themes and categories of students’ reflections Themes

Categories

Sample statements

Academic enhancement

Gained knowledge in materials production

“Conducting readability testing to the reading materials made the texts student-centered”

Personal growth

Applied theories in real context “I learned to meticulously find the right vocabulary to make effective stories”

Civic engagement

Developed self-awareness

“It made me a better person and educator. I recognized my strengths and the areas I needed to boost my skills to become better”

Developed compassion

“I have high respect to these teachers who balanced demanding actual remote teaching and conducting remedial program for their students”

Valued volunteerism despite the “In the future, people might need pandemic more helping hands, and I am willing to lend mine”

were in the reading frustration level. We applied the learnings that we have extracted from our class.” Despite the challenges during the pandemic, the students highlighted that they enjoyed learning about materials development. Student C wrote, “We really enjoyed this class because it can help us in our future endeavors. We were tasked to create learning materials for the action research that can help the students of Olango Island. It was learning with a purpose.” In addition, Student D said, “During the process, we applied all our learnings in materials development.” 4.2 Project Developed Personal Growth and Community Engagement Another positive impact of this e-SL was developing students’ personal growth and public service. The students wrote that the experience developed their self-awareness, collaboration, compassion, and fulfillment. Student F wrote, “This made me a better person and educator. I have recognized my strengths.” In addition, the experience also promoted students’ collaborative skills. Student H said, “I developed my skills to communicate effectively with other professionals.” Student I added that the experience allowed her to collaborate with others who have different expertise. “The project developed compassion and fulfillment among the students. Student G said, “We have come to be more compassionate in helping others.” For fulfillment, Student K said, “The appreciation we got from designing the layout of the materials made us excited to produce more.” In the FGD, the students claimed

28

A. Vilbar

that the e-SL project helped them in improving their mental health. They shared that meeting new teachers and having purposeful engagements eased their loneliness during the lockdowns. 4.3 Community Developed Research Skills All data revealed that the project developed the teachers’ skills in conducting action research and developing their instructional materials. The teachers acknowledged the significant role of the project in incapacitating their skills. From their reflections, they described their coaches as their source of knowledge and inspiration. Teacher 1 said, “The crucial part was designing the reading materials. Our coaches did their best to provide original samples.” The coaches guided them in their curriculum and materials development. Teacher 2 said, “Having coaches is truly a blessing. They were able to give us advice in finishing our remedial materials.” Teacher 3 shared, “My coaches taught me in using readability test as a scientific way to gauge the reading level of the text. It made my reading materials suited to my student’s level.” Grateful for the new online teaching website, Teacher 4 said, “My coach introduced me to Canva in lay-outing my lessons.” The teachers’ ability to produce effective instructional materials was validated by independent experts who assessed their outputs in the Expert’s Evaluation stage. This stage aims to promote external quality assurance of the material’s instructional design, accuracy of information, and cultural appropriacy [29]. In this stage, the evaluators provided constructive criticisms to revise their materials. Each group had three independent evaluators who had at least a master’s degree in education and had substantial teaching experience. The evaluation showed that all materials received a “Passed” rating in their content, format, presentation, and accuracy of information. It proved that the e-SL project helped nurture the public-school teachers to become action researchers and materials developers. Figure 1 shows an example of an original short story written by one of the volunteer coaches Ms. Gina Mantua-Panes. The short story was used in teaching reading comprehension. The illustrations were from Canva. This pilot testing aimed to provide empirical support regarding the experience of the end-users of the proposed reading and mathematics materials [29]. Findings showed that their students claimed that the reading texts were interesting and informative. Their students said they enjoyed reading the stories because of their topics, layout, and colored illustrations. Moreover, the students who used the math videos described the episode on Integers as interactive and easy-to-understand. All data substantiate that the e-SL project promoted academic enhancement, community engagement, and personal growth even without the FtF components. It developed the graduate students’ understanding of materials production by applying these skills to the community. The students highlighted that they needed to master the principles in materials design because they needed to help the community. As Student E’s reflection inscribed, “We were tasked to coach teachers on how to design effective remedial materials. It was learning with purpose.” The findings prove that e-SL is an effective pedagogy for it offers a unique transfer of learning in authentic contexts [8, 22]. e-SL’s empowered the students to appreciate

e-SL to Sustain Instruction with Civic Engagement

29

Fig. 1. An original short story to develop reading comprehension.

the course despite the health crisis, which led to a cathartic feeling of fulfillment [9, 14, 19, 20, 30]. Consequently, this fulfillment of helping the Olango teachers improved their mental health during the lockdowns. Having meaningful coaching with the teachers diverted their feeling of isolation into being productive. Their testimonies prove that similar to FtF-SL, e-SL has a transformative nature that can facilitate knowledge and volunteerism [5]. The learning of the course was directed on the needs of the community. 4.4 Technologies Sustained the Meaningful Collaboration Another significant factor in sustaining the project was utilizing various online and offline collaboration platforms like Zoom, Facebook Messenger, and Google Meets. Student A said, “The online Zoom Team Building helped us in making rapport with the Olango Island teachers.” Student B added, “Zoom and Google Meets were not my daily teaching tools before, but now, they have become my new normal platforms in meeting the teachers.” When asked to distinguish the platforms, all participants said they used Zoom for collaboration when the task required critical discussion. These tasks included presenting the syllabus, analyzing the content. They added that they preferred Zoom because it had more advanced features like having Breakout rooms and user-friendly file-sharing procedures. They used Google Drive to store and edit files. However, they used Facebook Messenger when they needed urgent replies. The data show that despite being a novice in using the digital platforms, all participants successfully navigated these technologies in their fashion to sustain the project. Admittedly, they shared initial technical difficulties in using the platforms but later discovered the affordances of Zoom. However, these issues did not dampen their collaboration. They agreed to collaborate at night when the connectivity was more stable. These adversities were transformed into compassion and action. The experience prove

30

A. Vilbar

that internet connectivity and technology are integral to sustaining e-SL and maintaining the emotional connections between the volunteers and community to facilitate the transfer of learning [22].

5 Conclusion The service learners’ experience promoted content course, personal growth, and community engagement even without FtF interaction during the pandemic. e-SL motivated the graduate students to learn materials production with a purpose. Their purpose was to help the community with their action research remedial program. Consequently, the project developed the community teachers’ skills in conducting action research to improve their students’ reading comprehension and mathematical skills. The data offered optimism to studies that provided constructive warnings to e-SL implementers on creating meaningful online communications during the pandemic [3, 11]. In this study, there were struggles in the initial online collaboration. However, the appropriate use of telecollaboration platforms such as Zoom, or social media still developed sincerity and sustained meaningful collaboration despite the absence of FtF engagements. The participants confirmed that online collaboration promoted better time management, participation, and convenience. These platforms can be the alternative or the mainstream tools for community engagements in a post- COVID-19 pandemic context. These online tools have features to simplify human communication without jeopardizing the safety of the volunteers and the community. As a post-COVID-19 pedagogy, service-learning can be redesigned into a hybrid SL model to respond to the need for authentic face-to-face instruction while still using the remote learning strategy of telecollaboration. Hybrid SL is a type of service-learning which uses some aspect of teaching and/or service conducted online or FtF [21]. The instruction may be executed online while the service is onsite or inversely [30]. Choosing the modality must be consultative depending upon the context of the course, the students, the community’s connectivity, and health protocols. A hybrid SL could happen when classes are conducted in FtF; the students would serve a community using telecollaboration platforms. The students could conduct online tutorials for the community. Another possibility could be when classes are conducted online; then, students would serve a community onsite. The students could start their service online and have a culmination in the community. The use of hybrid SL can promote the project’s sustainability, cost-efficiency, and marginal virus transmission risks to the students and the community. Acknowledgment. I would like to acknowledge the University of the Philippines Cebu, Central Visayas Studies Center, UP Cebu Master of Education Program, UP Cebu Research Ethics Committee, and UP Cebu Ugnayan ng Pahinungod for the support.

e-SL to Sustain Instruction with Civic Engagement

31

References 1. UNESCO: 1.37 billion students now home as COVID-19 school closures expand, ministers scale up multimedia approaches to ensure learning continuity (2020). https://en.unesco.org/ news/137-billion-students-now-home-covid-19-school-closures-expand-ministers-scalemultimedia 2. Hodges, C., Moore, S., Lockee, B., Trust, T., Bond, A.: The Difference Between Emergency Remote Teaching and Online Learning (2020). https://er.educause.edu/articles/2020/3/thedifference-between-emergency-remote-teaching-and3. Mejia, A.: “Plan for the worst, hope for the best, but realistically, expect a combination of both”: lessons and best practices emerging from community-engaged teaching during a health crisis. J. High. Educ. Outreach Engagem. 25(3), 35–50 (2021) 4. Bringle, R.G., Hatcher, J.A.: A service-learning curriculum for faculty. Mich. J. Community Serv. 2, 112–122 (1995). https://www.semanticscholar.org/paper/A-Service-Learning-Curric ulum-for-Faculty-Bringle-Hatcher/c5c9440c4566d1fe7b78b9a7f57bf5e509926f48 5. Playford, D., et al.: Twelve tips for implementing effective service learning. Med. Teach. 41(1), 24–27 (2019). https://doi.org/10.1080/0142159X.2017.1401217 6. Wurr, A.: Advances in service-learning research with English language learners. J. Serv. High. Educ. 8 (2018) 7. Malvey, D.M., Hamby, E.F., Fottler, M.D.: E-service learning: a pedagogic innovation for healthcare management education. J. Health Adm. Educ. 23(2), 181–198 (2006) 8. Andrews, U., Tarasenko, Y., Holland, Y.: Complexities of Coordinating Service-Learning Experiences in Rural Communities, pp. 192–210 (2020). https://doi.org/10.4018/978-1-79983285-0.ch012 9. Schmidt, M.E.: Scholarship of Teaching and Learning in Psychology Embracing e-Service Learning in the Age of COVID and Beyond Embracing e-Service Learning in the Age of COVID and Beyond (2021) 10. Lewis, S.M., Strano-Paul, L.A.: A COVID service-learning initiative: emotional support calls for the geriatric population. J. Am. Geriatr. Soc. 69(2), E4–E5 (2021). https://doi.org/10.1111/ jgs.17003 11. Guy, B., Arthur, B.: Impact of COVID-19 on a participatory action research project: grouplevel assessments with undergraduate women in engineering. J. High. Educ. Outreach Engagem. 25(3), 5–14 (2021) 12. Flores, D.D., Bocage, C., Devlin, S., Miller, M., Savarino, A., Lipman, T.H.: When community immersion becomes distance learning: lessons learned from a disrupted semester. Pedagog. Heal. Promot. 7(1), 46–50 (2021). https://doi.org/10.1177/2373379920963596 13. Grenier, L., Robinson, E., Harkins, D.: Service-learning in the COVID19 era: Learning in the midst of crisis. Pedagog. Hum. Sci. 7(1), 5 (2020) 14. Tian, Q., Noel, Jr. T.: Service-learning in catholic higher education and alternative approaches facing the COVID-19 Pandemic. J. Cathol. Educ. 23(1), 184–196 (2020). https://doi.org/10. 15365/joce.2302142020 15. Meuser, T., Cohen Konrad, S., Robnett, R., Brooks, F.: Telecollaboration in gerontology service learning: Addressing isolation and loneliness in a pandemic. Gerontol. Geriatr. Educ. 1–16 (2021).https://doi.org/10.1080/02701960.2021.1956489 16. Seru, E.: Critical, interdisciplinary, and collaborative approaches to virtual communityengaged learning during the COVID-19 pandemic and social unrest in the twin cities. J. High. Educ. Outreach Engagem. 25(3), 79–90 (2021) 17. Bringle, R.G., Clayton, P.H.: Integrating service learning and digital technologies: examining the challenge and the promise. RIED. Rev. Iberoam. Educ. Distancia 1, 43–65 (2020). https:// doi.org/10.5944/ried.23.1.25386

32

A. Vilbar

18. Leung, H., Shek, D.T.L., Dou, D: Evaluation of service-learning in project we can under COVID-19 in a Chinese context. Int. J. Environ. Res. Public Health 18(7) (2021). https://doi. org/10.3390/ijerph18073596 19. Ley, C., Angert, D., Hudelson, T.: Adapting global service-learning project and community partnership outcomes using a ‘tele-engineering’ approach in response to the COVID-19 pandemic adapting global service- learning project and community partnership outcomes 8(1) (2021). https://doi.org/10.5703/1288284317408 20. Lin, L., Shek, D.T.L.: Serving children and adolescents in need during the COVID-19 pandemic: evaluation of service-learning subjects with and without face-to-face interaction. Int. J. Environ. Res. Public Health 18(4), 1–16 (2021). https://doi.org/10.3390/ijerph18042114 21. Waldner, L., McGorry, S., Widener, M.: Extreme e-service learning (XE- SL): E-service learning in the 100% online course. J. Online Learn. Teach. 6(4), 839 (2010) 22. Stefaniak, J.: A systems view of supporting the transfer of learning through E-service-learning experiences in real-world contexts. TechTrends 64(4), 561–569 (2020). https://doi.org/10. 1007/s11528-020-00487-3 23. Dancel, R.: Cebu City looms as Philippines’ new COVID-19 epicentre. The Strait Times, 29 June 2020 24. Mason, P., Augustyn, M., Seakhoa-King, A.: Exploratory study in tourism: designing an initial, qualitative phase of sequenced, mixed methods research. Int. J. Tour. Res. 12(5), 432–448 (2010). https://doi.org/10.1002/jtr.763 25. Mitchell, T.D.: Using a critical service-learning approach to facilitate civic identity development. Theory Pract. 54(1), 20–28 (2015). https://doi.org/10.1080/00405841.2015.977657 26. Liu, Y.: Using reflections and questioning to engage and challenge online graduate learners in education. Res. Pract. Technol. Enhanc. Learn. 14(1), 1 (2019). https://doi.org/10.1186/ s41039-019-0098-z 27. Harding, J.: Identifying themes and coding interview data: reflective practice in higher education. Identifying Themes Coding Interview Data Reflective Pract. High. Educ. (2015). https:// doi.org/10.4135/9781473942189 28. Ash, S.L., Clayton, P.H.: The articulated learning: an approach to guided reflection and assessment. Innov. High. Educ. 29(2), 137–154 (2004). https://doi.org/10.1023/B:IHIE.000 0048795.84634.4a 29. Lin, Y.K., et al.: Development and pilot testing of an informed consent video for patients with limb trauma prior to debridement surgery using a modified Delphi technique. BMC Med. Ethics 18(1), 1–12 (2017). https://doi.org/10.1186/s12910-017-0228-3 30. Dovi, K., Chiarelli, J., Franco, J.: Service-learning for a multiple learning modality environment. J. Chem. Educ. 98(6), 2005–2011 (2021). https://doi.org/10.1021/acs.jchemed.0c0 1475

Evaluating E-Learning Process on Virtual Classroom Systems Using an ISO-Based Model Nicholas Coulianos , Athanasia Sapalidou , Akrivi Krouska(B) Christos Troussas , and Cleo Sgouropoulou

,

Department of Informatics and Computer Engineering, University of West Attica, Ag. Spyridonos Street, 12243 Egaleo, Greece {ice18390113,ice18390163,akrouska,ctrouss,csgouro}@uniwa.gr

Abstract. E-learning is the process of acquiring knowledge through electronic devices such as computers and smart devices connected to the internet, facilitating users to learn anytime and anywhere. E-learning through modern virtual classroom platforms combines a variety of technological tools aiming at successful and effective distance learning, enhancing the sense of social presence and team spirit among students, while allowing teachers to communicate with students in ways that traditional in-person learning does not offer. This article addresses the issue of evaluating e-learning systems that utilize the virtual classroom model, highlighting the features and requirements they must meet. As such, an ISO-based standard for the evaluation of e-learning platforms in a virtual classroom is proposed, aiming at the satisfaction of students and teachers. Finally, four modern software products that are widely used for the e-learning process in a virtual classroom are assessed, based on the proposed evaluation standard. Keywords: E-learning · Evaluation · ISO 25010 · Virtual classroom

1 Introduction The term “e-learning” refers to the process of acquiring knowledge through an online or computer education system [1, 2]. It is a complete, structured and integrated learning experience provided to student users electronically, via electronic devices such as computers and mobile phones connected to the internet, without restrictions of time and place. According to Chou [3], e-learning is characterized as an “information delivery tool” that operates via the Internet synchronously or asynchronously depending on the desire of users. Al-Nuaim [4] used the term “virtual classroom” to describe the place where the online educational process allows students and teachers to synchronize communication and exchange opinion through text, audio, videos, interactive whiteboards, and other similar features and facilities in the same way that they would interact face-to-face in the real classroom. Virtual classroom can complement and substitute the traditional in-person learning process, offering all the necessary tools so that both teachers and students are consistent with their responsibilities, increasing the active role of students © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 33–45, 2023. https://doi.org/10.1007/978-3-031-17601-2_4

34

N. Coulianos et al.

and enhancing their involvement with the educational subject [5]. However, there are challenges in choosing between the numerous virtual classroom software products that are available in the market [6]. It is necessary to ensure that these tools will provide rich services meeting the users’ demands. This article presents the main benefits of using the virtual classroom model for learning. The research is addressed to educators, researchers and programmers, and focuses on the evaluation of e-learning systems that implement the virtual classroom aiming to highlight the features and requirements that must be met. We propose a methodical approach by identifying the basic functions that e-learning platforms should provide in a virtual classroom, in order to meet the needs of both students and teachers. Finally, we demonstrate and evaluate four modern software products that are widely used for promoting the e-learning process through the virtual classroom model.

2 Advantages of E-learning Systems and Platforms E-learning platforms come to complement and expand in-person learning, offering all the necessary tools to face the difficulties and problems of face-to-face instruction, some of which are the limited learning material and the technologies that can be used, the limitation of time and place of the lesson, the tough communication between students and teachers and the collaboration between the students [7–9]. As indicated by Ouadoud et al. [6] in their research, the characteristics of e-learning platforms are: 1. They can extend the education through the additional material that students can access via the educational platform. This material can be interactive contributing to a better and deeper understanding of the educational subject [10]. 2. They can be characterized by flexibility, as they offer freedom in time and place of attendance, which are very often negotiable between students and teachers [11], thus allowing students to work and have an active social life [12]. 3. In addition to this, they allow those who find educational “obstacles” to have equal opportunities to learning, overcoming them. 4. They introduce new teaching practices owing to the use of innovative technologies that facilitate and encourage communication, aiming at the most effective and efficient education. 5. As a result, they render collaborative projects easier to manage [13]. 6. In such a learning environment, where communication and interaction between students as well as between students and teachers is constant, the students are able to produce material, process and document existing material, as well as share their knowledge and observations with other students contributing to self-evaluation and self-improvement [2]. One of e-learning’s biggest innovations and achievements is the virtual classroom, which is designed in order to “transfer” the traditional classroom on the Internet, allowing and enhancing remote access for all users, immediate interaction among students and educators, as well as communication and the ability of collaboration among students

Evaluating E-Learning Process on Virtual Classroom Systems

35

[14]. However, it differs from the traditional classroom by focusing on the importance of communication between students, rather than the “teacher-student” interaction aiming at educational success [15]. In order to highlight the advantages of using modern virtual classroom systems in education, it is necessary to be considered in comparison to two axes: • the asynchronous e-learning systems, and • the real in-person classroom. Firstly, concerning the advantages of the modern virtual classroom compared to asynchronous e-learning systems, the following can be observed: 1. The combination of real time audio, visual and written interaction render it more attractive and enjoyable compared to asynchronous e-learning methods [16]. 2. At the same time, within the collaborative learning environment of the virtual classroom where interaction among students is constantly encouraged, it is easier to achieve educational goals, as students feel they are part of a team, having common motivation and aspirations and setting common goals they work for to perform [17]. On the other hand, concerning the advantages of the modern virtual classroom compared to the real in-person classroom, it should be noted that: 1. Virtual classroom greatly reduces the stress that students may face when it is necessary to actively participate in face-to-face discussions with their classmates or teachers [16]. 2. Virtual classroom platforms are free from size constraints (i.e., the number of participants) as well as the features and services they provide. Therefore, they can serve multiple users - both students and teachers - regardless of their educational level, age or special needs [14], contributing to lifelong learning and training.

3 Software Evaluation Standards and Models Providing high-quality software and computer systems is of great importance. Software products concern both those who develop them and those who use them whether they are operators or customers [18]; in case of e-learning systems, the users can be the educational institutions, teachers and students. According to Dragogiannis et al. [19], the assessment of an information system is considered to be a fundamental process in order to select the most appropriate and effective system that will fully meet the needs and desires of all of the above. In order to evaluate the quality and reliability of a software product, it is necessary to completely specify all the features and requirements that it must meet [18]. The evaluation of a software product can be achieved by using standards that determine the necessary quality features of each system. When developing software products, the use of evaluation standards helps to identify:

36

N. Coulianos et al.

1. The requirements and quality characteristics of the software system. 2. The quality control criteria that ensure the quality of the software system. 3. The criteria by which a software product is considered acceptable [18]. Every system needs to provide users with quality assurance and reliability. Using standards helps systems to be secure, reliable and high quality, ensuring that they comply with consumer demands. 3.1 The ISO 25010 Evaluation Standard ISO 25010 is a well-known and accepted quality standard that has been used to evaluate numerous software products [22]. This evaluation model repealed and replaced ISO/IEC 9126-1: 2001, focusing on compatibility and security, which until then used to be subfeatures of the functionality feature. The ISO 25010 quality assessment model consists of eight quality features: functionality, efficiency, compatibility, usability, reliability, safety, maintainability and portability. For each of the above characteristics, individual sub-characteristics are defined. Figure 1 summarizes all the characteristic and sub-characteristic of the ISO 25010.

Fig. 1. ISO 25010.

Studying the literature on the evaluation of e-learning systems, there are many researchers who modified ISO 25010 in terms of the evaluation criteria, according to their systems’ needs. Al-sarrayrih et al. [20] utilized the previous version of ISO 25010, the ISO 126, along with the users’ opinions and preferences, for the evaluation of a learning management system used at the Berlin Institute of Technology. Respectively, Balahadia and Urera [21], based on the characteristics of ISO 25010, conduct sampling using a questionnaire to find the most important quality characteristics of an e-learning system with the aim of creating an educational platform. Furthermore, Manglapuz and Lacatan [22] evaluate academic management android applications using the ISO 25010 model, aiming at the development of new e-learning

Evaluating E-Learning Process on Virtual Classroom Systems

37

software as well as the establishment of new techniques and tools for the evaluation of e-learning systems. These two factors will determine the academic progress of students. In Krouska et al. [23], the evaluation of delivering e-learning through social media platforms is based on the quality characteristics of the ISO 25010 standard, while adding some additional specific quality sub-characteristics, which they considered to be directly related to e-learning and social media, categorized per characteristic of the ISO 25010 standard. Aguirre et al. [24] rely on the ISO 25010 in order to define the attributes that should be taken into consideration to achieve user satisfaction when using e-learning systems: utility, trust, pleasure and comfort.

4 Evaluation of E-learning Systems in Virtual Classroom 4.1 Introducing an Adjusted Quality Evaluation Model For the evaluation of virtual classroom e-learning systems, the ISO 25010 standard was used, adapting it properly based on the requirements of the field. As such, twenty-four quality sub-characteristics were incorporated in this standard, creating an ISO-based evaluation model for virtual classroom systems. More specifically, the characteristics that were added are the following: • Ten sub-characteristics related to functional sustainability: Dedication Dashboard, Text Chat, Voice Chat (VOIP), Video Chat (two-way), Break Room, Hand Raising, Voting, Polls, Tests and Questionnaires • One performance efficiency-related sub-characteristic: Asynchronous browsing • Eight compatibility sub-characteristics: Interactive Whiteboard, PowerPoint Presentation, Multimedia Presentation, Application Sharing and Integration, Record and Replay, File Sharing, and Content Library. • Four usability sub-characteristics: User manual, Layouts and Templates, Accessible videos, and Accessible documents • One security-related sub-characteristic: Password Security • One sub-characteristic related to maintainability: Class evaluation • Two sub-characteristics related to portability: Cross platform, Plugins requirements Table 1 illustrates all the characteristic and sub-characteristic of the proposed evaluation model. 4.2 Evaluation Results Four modern virtual classroom software widely used in the e-learning process, namely Adobe Connect, Blackboard Collaborate, LearnCube and VEDAMO Virtual Classroom, were evaluated based on the proposed model. Adobe Connect users are almost equally divided into small, medium and large businesses (with small businesses leading by the small ones by less than 1% of all users). In addition, Blackboard Collaborate is used extensively by large companies (with more

38

N. Coulianos et al. Table 1. Characteristics of the proposed evaluation model.

ISO 25010 basic features

ISO 25010 individual characteristics

Special quality sub-characteristics related to virtual classrooms

Functional sustainability

Functional Completeness

Dedication dashboard Text chat Voice chat Video chat Hand Raising Break Room Voting Polls Tests Questionnaires

Performance Efficiency

Resource utilization

Asynchronous browsing

Compatibility

Co-existence

Interactive Whiteboard PowerPoint Presentation Multimedia Presentation

Interoperability

Record and Replay (voice, text and screen) File Sharing Content Library Application Sharing Integration

Usability

Operability

User manual

User Interface Aesthetics

Layouts and Templates

Accessibility

Accessible videos

Security

Confidentiality

Password security

Maintainability

Modifiability

Class evaluation

Portability

Adaptability

Cross platform

Installability

Plugins requirements

Accessible documents

than 1000 users). Finally, as far as LearnCube and VEDAMO Virtual Classroom are

Evaluating E-Learning Process on Virtual Classroom Systems

39

concerned, they are mainly used by small businesses (with less than 50 users), with large companies not choosing them almost none at all (0% and 2% respectively)1 . Functional Sustainability According to ISO 25010, functional sustainability includes the characteristic of functional completeness from which accrue the special quality sub-characteristics Dedication Dashboard, Text Chat, Voice Chat (VOIP), Video Chat (two-way), Break room, Hand Raising, Voting, Polls, Tests and Questionnaires that are essential for virtual classroom e-learning systems. The Dedication Board refers to a table that contains statistics about the courses, the lectures etc. that students attended the most. Text Chat, Voice Chat (VOIP) and Video Chat (two-way) concern the communication between students, as well as between students and teachers, through text and voice messages, phone calls and video calls. Hand raising is a function that allows students to request to speak during a session. Break room is a separate “room” in which the teacher/organizer of a session can “move” the participants temporarily. Voting, Polls, Tests and Questionnaires are functions that allow students to self-evaluate, evaluate the educational process and make decisions together with their classmates and teachers. Out of all the evaluated platforms, only Adobe Connect provides the Dedication Board and Poll feature. Whereas, all the platforms provide Text Chat, Voice Chat (VOIP), Video Chat (two-way), Hand raising, Voting, Tests and Questionnaires. Performance Efficiency As shown in Fig. 1, the performance efficiency includes the resource utilization characteristic from which accrue the special quality sub-characteristic Asynchronous browsing related to virtual classroom e-learning systems. Asynchronous browsing is the ability of users to use the software outside of the session/call, accessing its other functions. The only platform that provides this feature is Adobe Connect. Compatibility Compatibility includes the characteristics of co-existence and interoperability. As such, the following special quality sub-characteristics are desirable to be incorporated in virtual classroom e-learning systems: 1. Coexistence has the special sub-characteristics Interactive Whiteboard, PowerPoint Presentation and Multimedia Presentation. The Interactive Whiteboard is an electronic whiteboard on which participants in a session / call can write in real time. In addition, the PowerPoint Presentation and other multimedia presentation concern the user’s ability to share PowerPoint presentations as well as various multimedia such as photos, videos, etc.

1 G2: Simple comparison among software and users’ rating, www.g2.com/compare/adobe-

connect-vs-blackboard-collaborate-vs-learncube-vs-vedamo-virtual-classroom, last accessed 2021/31/12.

40

N. Coulianos et al.

2. Interoperability includes the special sub-characteristics Application Sharing, Integration, Record and Replay, File Sharing, and Content Library. Application Sharing refers to the user’s ability to share a third-party application through the software. Additionally, if during a session the software, through its interface, informed the user that an application is already embedded and can use it, then we would be talking about the function of Integrations. The Record and Replay function means screen and / or audio recording during a session. The recorded session will be available for playback afterwards. File Sharing is the user’s ability to send and receive files to and from other software users. Finally, the Content Library stores all the files and data that are necessary for the lessons that take place. Whiteboard, PowerPoint Presentation, Multimedia Presentation, Application Sharing, Record and Replay and Content Library are provided by all four software. However, the Integration feature is only found in Adobe Connect, while the File Sharing feature is only found in Adobe Connect and the VEDAMO Virtual Classroom. Usability Figure 1 shows that usability includes the characteristics of operability, user interface aesthetics and accessibility. The special quality sub-characteristics related to virtual classroom e-learning systems for each characteristic are as follows: 1. Operability includes the special sub-characteristic User manual. User manual is a user-friendly manual to guide and facilitate the users through the system. 2. The special sub-characteristics Layouts and Templates pertain to the user interface aesthetics. Layouts and Templates allow the user to customize the screen and interface, aiming at better organization and attractiveness. 3. Accessibility has the special sub-characteristics Accessible Videos and Accessible Documents. Accessible videos and documents are aimed at users with visual and hearing impairments, in order to participate actively, continuously and without disturbance in the educational process. User manual and Accessible Video features are provided by all four software. However, the Accessible Documents function is not provided by VEDAMO Virtual Classroom, while Layouts and Templates are only available on Adobe Connect and VEDAMO Virtual Classroom. Security Similarly, according to ISO 25010, security includes the characteristic of confidentiality in which the special quality sub-characteristic Password Security has been added as required feature for virtual classroom e-learning systems.

Evaluating E-Learning Process on Virtual Classroom Systems

41

Password Security refers to a user password that each user is required to have in order to access the software. All four software provide this function. Maintainability As shown in Fig. 1, maintainability includes the modifiability characteristic from which emerges the special quality sub-characteristic Class evaluation that is essential for virtual classroom e-learning systems. Class evaluation offers students the opportunity to assess the educational process. Out of all the four platforms, only LearnCube provides this feature. Portability Finally, portability includes the characteristics of adaptability and installability. The following special quality sub-characteristics related to virtual classroom e-learning systems corresponds to each characteristic: 1. Adaptability includes the special Cross platform sub-feature. The Cross platform feature means that the software is available for installation and use on different operating systems (e.g. Windows, Linux, Mac). 2. Installability has the special sub-feature Plugins requirements. A Plugins requirement means that in order for the software to work, a plugin (e.g., Java) must be installed on the computer. In general, we could say that it is preferable for software not to have plugins requirements. None of the four platforms has plugins requirements. It is apparent that among the four software products evaluated, Adobe Connect outperforms the other three; as it is the only one that provides features that no other platform does, such as the Dedication dashboard, Polls, Asynchronous browsing, Integration and File sharing. These features are very likely to motivate those looking for a virtual classroom application (teachers, students, educational institutions) to choose it (Adobe Connect). Tables 2 and 3 summarize the results of the software evaluation based on the adjusted proposed standard. More specifically, Table 2 shows all the quality features we searched for, as well as which of the four products have them. Table 3 records how many special quality features per ISO characteristic each software has, while in the last two lines the total number of available features is counted and the final score is converted to a percentage.

42

N. Coulianos et al. Table 2. Evaluation of the software products using the proposed standard.

Evaluating E-Learning Process on Virtual Classroom Systems

43

Table 3. Rating the software products based on the evaluation results. ISO 25010 features

Satisfaction of special requirements Adobe Connect

Blackboard Collaborate

LearnCube

VEDAMO

Functional sustainability

10/10

8/10

8/10

8/10

Performance Efficiency

1/1

0/1

0/1

0/1

Compatibility

8/8

6/8

6/8

7/8

Usability

4/4

3/4

3/4

3/4

Security

1/1

1/1

1/1

1/1

Maintainability

0/1

0/1

1/1

0/1

Portability

1/2

1/2

1/2

1/2

Total score

25/27

19/27

20/27

20/27

Score in percentage

92.59%

70.37%

74.07%

74.07%

5 Conclusion Summarizing, this article addresses the issue of e-learning systems evaluation and more specifically, e-learning systems that utilize the virtual classroom model. Regarding the main benefits of using e-learning systems, they can be the rich educational material that students can access, the use of new technologies that encourage communication and collaboration, as well as the freedom in time and place of learning process. In particular for virtual classrooms, another important benefit is that they can serve a large number of users regardless of their educational level, age or special needs. A conclusion emerged from the current research is that each e-learning platform for virtual classroom is unique, provides different technological tools and focuses on different pedagogical approaches. Similarly, each user of such a platform, whether is a student, teacher or educational institution, has different needs and demands (from a platform). Therefore, it is not always easy to define a typical model for evaluating the quality of such systems. However, there are guidelines that can be followed to formulate an evaluation model based on the features and functions that are considered the most important. For this reason, an ISO-based standard for the evaluation of e-learning platforms in a virtual classroom is proposed, aiming at the satisfaction of students and teachers. The success of e-learning depends on the continuous evaluation of the available software products. Both educators and developers need to understand how to make the most of all the tools provided by e-learning technology in order to improve the educational process.

44

N. Coulianos et al.

References 1. Al-Rahmi, W.M., Othman, M.S., Yusuf, L.M.: Exploring the factors that affect student satisfaction through using e-learning in Malaysian higher education institutions. Mediterran. J. Soc. Sci. 6(4) (2015) 2. Krouska, A., Troussas, C., Sgouropoulou, C.: A cognitive diagnostic module based on the repair theory for a personalized user experience in e-learning software. Computers 10, 140 (2021). https://doi.org/10.3390/computers10110140 3. Chou, C.: Interactivity and interactive functions in web-based learning systems: a technical framework for designers. Br. J. Edu. Technol. 34(3), 265–279 (2003) 4. Al-Nuaim, H.A.: The use of virtual classrooms in e-learning: a case study in King Abdulaziz University, Saudi Arabia. E-Learn. Dig. Media 9(2), 211–222 (2012) 5. Wardani, R., et al.: Improving online learning interactivity with 3D virtual classroom models. J. Phys. Conf. Ser. 2111(1), 10 (2021) 6. Ouadoud, M., Chkouri, M.Y., Nejjari, A., El Kadiri, K.E.: Studying and analyzing the evaluation dimensions of e-learning platforms relying on a software engineering approach. Int. J. Emerg. Technol. Learn. 11(01), 11–20 (2016) 7. Troussas, C., Krouska, A., Alepis, E., Virvou, M.: Intelligent and adaptive tutoring through a social network for higher education. New Rev. Hypermedia Multimedia 26(3–4), 138–167 (2020) 8. Troussas, C., Krouska, A., Giannakas, F., Sgouropoulou, C., Voyiatzis, I.: An alternative educational tool through interactive software over Facebook in the era of COVID-19. In: Novelties in Intelligent Digital Systems, pp. 3–11. IOS Press (2021) 9. Troussas, C., Krouska, A., Sgouropoulou, C.: Enriching mobile learning software with interactive activities and motivational feedback for advancing users’ high-level cognitive skills. Computers 11(2), 18 (2022) 10. Troussas, C., Krouska, A., Sgouropoulou, C.: A novel teaching strategy through adaptive learning activities for computer programming. IEEE Trans. Educ. 64(2), 103–109 (2020) 11. Suryani, N.K., Sugianingrat, W.: Student e-learning satisfaction during the covid-19 pandemic in Bali, Indonesia - Kepuasan Siswa Mengikuti E-Learning Selama Pandemi Covid-19 di Indonesia. Jurnal Economia 17(1), 141–151 (2021) 12. Moore, M.G., Kearsley, G.: Distance Education: A Systems View, 2nd edn. Wadsworth, Belmont, CA (2005) 13. Krouska, A., Virvou, M.: An enhanced genetic algorithm for heterogeneous group formation based on multi-characteristics in social-networking-based learning. IEEE Trans. Learn. Technol. 13(3), 465–476 (2019) 14. Adewale, O.S., Ibam, E.O., Alese, B.K.: A web-based virtual classroom system model. Turkish Online J. Dist. Educ. 13(1), 211–223 (2012) 15. Johnson, D.W.: Student-student interaction: the neglected variable in education. Educ. Res. 10(1), 5–10 (1981) 16. Daly, D., Rasmussen, A.V., Dalsgaard, A.: Learning about midwifery in another country from a distance: evaluation of a virtual classroom learning session. Nurse Educ. Today 75, 47–52 (2019) 17. Hiltz, S.R.: Correlates of learning in a virtual classroom. Int. J. Man Mach. Stud. 39, 71–98 (1993) 18. I.S.O.: Iso. IEC25010: 2011 Systems and software engineering–Systems and software Quality Requirements and Evaluation (SQuaRE)–System and software quality models. International Organization for Standardization (2011) 19. Dragogiannis, K., Papadopoulou, P., Pantelopoulou, S., Tsolou, O.: E-learning system evaluation for in-service education and training. In: 1st WSEAS International Conference on Computer Supported Education (COSUE 2013). Athens (2013)

Evaluating E-Learning Process on Virtual Classroom Systems

45

20. Al-Sarrayrih, H., Knipping, L., Zorn, E.: Evaluation of a MOODLE based learning management system applied at berlin institute of techology based on ISO-9126. In: Conference ICL2010. Belgium (2010) 21. Balahadia, F.F., Urera Jr., F.L.: ICTeachMUPO: an evaluation of information e-learning module system for faculty and students. Int. J. Comput. Sci. Res. 3(1), 163–188 (2020) 22. Manglapuz, S.J.R., Lacatan, L.L.: Academic management android application for student performance analytics: a comprehensive evaluation using Iso 25010:2011. Int. J. Innov. Technol. Explor. Eng. 8(12) (2019) 23. Krouska, A., Troussas, C., Virvou, M.: A literature review of social networking- based learning systems using a novel ISO-based framework. Intell. Dec. Technol. 13(2), 1–17 (2019) 24. Aguirre, A.F., Villareal-Freire, Á., Gil, R., Collazos, C.A.: Extending the concept of user satisfaction in e-learning systems from ISO/IEC 25010. In: Marcus, A., Wang, W. (eds.) DUXU 2017. LNCS, vol. 10290, pp. 167–179. Springer, Cham (2017). https://doi.org/10. 1007/978-3-319-58640-3_13

SERVE as Instructional Design for Low-Connectivity Online Self-directed Modules Jeraline Gumalal(B)

and Aurelio Vilbar

University of the Philippines Cebu, 6000 Cebu City, Philippines [email protected]

Abstract. COVID-9 pandemic caused suspensions of face-to-face classes and moved the teaching of science from physical to online. However, some high school students experienced unconducive home learning environment, economic issues, and internet connectivity limitations, thus making real time synchronous classes unfavorable and frustrating. This study aimed to address this issue by using the SERVE model to create and implement a self-directed earth science module anchored upon 4 stages – present, supplement, inquire, engage. It was found that the module promoted self-directed learning, learning in low connectivity situations, and science inquiry and reflective skills. Public schools undergoing remote learning modality may benefit from the SERVE model, however, the model should also be tested in a more massive scale. Keywords: Online learning · Self-directed module · Instructional design

1 Introduction The teaching of science in the Philippine high schools was disrupted by the COVID-19 pandemic thereby restricting the once experiential learning based course [1] to remote learning. This happened due to the suspension of face to face (f2f) classes, mandatory social distancing, and localized lockdowns [2, 3]. The situation favored the use of virtual learning environments (VLE) which allows schools to offer their courses online in either self-directed modules or synchronous real-time formats [4–6]. However, due to many students’ unconducive home learning environment, economic, and connectivity limitations, many cannot maximize real-time synchronous lectures [7]. They subsequently fail to turn in their requirements at the given time and affected their satisfaction, self-efficacy, and motivation to learn online [8, 9]. This research addressed the need of a state university high school to sustain its science remote learning in a low connectivity environment during the COVID-19 pandemic by designing an online self-directed module based on SERVE Model (SupplementEngage-Respond in a Virtual Environment). Although other scaffolded online learning instructional designs such as Mamun’s [10, 11] POEE (Predict, Observe, Explain, Evaluate) are available, SERVE is meant for a large low internet connectivity classroom that © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 46–51, 2023. https://doi.org/10.1007/978-3-031-17601-2_5

SERVE as Instructional Design

47

utilizes resource-based learning (RBL) and guided inquiry [12]. This model believes that student’s learning time, instructional tool, and space for self-inquiry can affect their online learning [8]. It posits that self-directed learning can be maximized through a good learner-content interaction [13, 14] and correct instructional planning. Studies has shown that online interactions and hybrid learning can be sustainable alternative to f2f instruction during the post-pandemic periods [15]. 1.1 SERVE, RBL, Guided Inquiry, Self-directed Learning The SERVE model focuses on the supplementation, inquiry, and engagement to support student’s self-direction through resource-based learning (RBL). Alongside scaffolded teaching instruction, RBL can enhance a learner’s self-directed learning through (1) clear learning objectives, (2) time management, (3) resources and responsive engagement, and (4) structured navigation with easy-to-understand instructions [16]. Table 1. The anticipated student workflow when developing a module using SERVE Stage

Anticipated asynchronous student workflow

Present

1. The student views objectives 2. The student views the 10-min main topic lecture video. The student is asked to either review when needed or take down notes 3. The student answers the 3-item conceptual questions and 5-item multiple choice questions on key terms

Supplement

1. The student views the curated video from an open education site. The video is usually 10-min long or less and must contain real life applications or situations related to the main topic 2. The student is encouraged to view or read suggested materials related to the video

Inquire

1. The student will be presented with 4-item conceptual questions scaffolded from the main lecture and supplementary video 2. The student is asked to create a cohesive essay from their answers to the questions 3. The student is given a rubric on the qualities needed from their outputs, a set deadline, and the output layout

Engage

1. The teacher gives feedback based on the rubric. The teacher delivers the feedback through the learning management system for the individual feedback. The student may communicate with the teacher regarding their work 2. The teacher gives class feedback noting on the concepts that needed to be reviewed and the skills that needs to be improved by the class. The delivery is through the social media chat group

This study investigated to what extent the online self-directed module anchored on SERVE model assisted the self-directed learning of seventy earth science students from a Philippine state university high school. The study utilized focus group discussion (FGD), performance rating from three independent raters, and open-ended survey to gather data.

48

J. Gumalal and A. Vilbar Table 2. The aspects of the self-directed module that the students find helpful

Aspects

FGD responses

Clear objectives, good lecture videos, and short lesson packages

“I can scan the whole lesson in minutes. The interface did not look overwhelming to study” “The instructions were very clear. I knew directly how important it is to watch the videos to have better insights about glaciers.” “All the videos are interesting and educational. I view them because I like them, and I need them for my writing output.”

Clear guide questions and exercise instructions

“The writing exercise is very achievable because of the guide questions. It allowed me to think deeply and look for reliable sources of information.” “The guide questions help me focus and allowed me to make my writing more organized and specific.”

Good alternative to real-time synchronous lectures

“Online synchronous classes can last for at least one hour which could consume so much internet data. Also, when it rained in my province, internet becomes unstable. But with asynchronous classes, I can access the module when the internet is reliable.” “I studied the module when I would be alone in the house. It could be in the morning or evening. I didn’t have my own room to study.”

Feedbacks

“The feedback made me more critical of my weaknesses in writing my task but made me feel more confident about my work.” “The feedback made my succeeding work better.”

2 Results The module utilized curated teacher-designed videos and open education videos for enriching science concepts in the module to develop student critical thinking [16]. Figure 1 shows the example lesson on Glaciers. The use of RBL in the online module develops motivation to learn, deeper knowledge, and self-directed learning [16, 17]. Students gave credit to their new learnings from the curated open educational videos within their required and suggested learning resources. Findings from the FGD (see Table 1) showed that the module promoted self-directed learning because it had a sound online instructional design in terms of clarity of objectives and scaffolded inquiry exercise, informative and interesting instructional videos for discussion and teacher feedback. All participants claimed the module’s graphical user interface was easy and easy to navigate.

SERVE as Instructional Design

49

Fig. 1. An example online module sub-lesson using the SERVE instructional stages

All participants were satisfied with the brief lesson package with achievable objectives. They also noted that the guided inquiry exercises are not stressful although cognitively demanding. The module provided additional references in case they wanted to know more about the topic and all participants admitted that they watched all the supplementary videos. Guided inquiry allows the students to reflect on their knowledge and encourages them to learn more [11]. The students’ testimonies proved that the module’s instructional design addressed the students’ needs and the course objectives. The students appreciated the module because it sustained their active learning without the demands of a synchronous class. This means that the module, which also follows the 4-step instructional design of Brame [18] for its videos, achieved proper signaling through thought-out text, segmenting of information for better knowledge digestion, weeding of unnecessary cognitive load, and matching with the correct modality. This also implies that the good learner-content interaction within the module allows students to have better learning engagements [14] (see Table 2). In terms of inquiry exercises, all students agreed their writing task was not a dreadful experience because of clear instructions and rubrics. Their testimonies were validated through the class’ inquiry writing exercise performance of 73.73% (Very good) by three independent raters. The rating criteria included use of scientific knowledge, reflective understanding of the concepts, clear and concise writing, and use of proper citation. Figure 2 is and informative/reflective answer from one of the students.

Fig. 2. This is a screenshot of Student G’s answer discussing that glacier can self-destruct (knowledge gained on Present stage) and can be indirectly destroyed by humans (reflection gained from Supplement stage).

The module is supportive of self-directed learning because it allows students to access the lessons whenever they felt they were in their best disposition to learn. The students

50

J. Gumalal and A. Vilbar

had metacognition in planning their learning. All participants highlighted that remote learning had caused anxiety, loneliness, and mental health issues [2, 7]. But using the module, they had the freedom to learn at their own pace. The students said that the teacher feedback helped them in assessing their performance and in improving their inquiry writing exercise outputs. All participants appreciated the overall feedback of their outputs in the social media group chat and the individualized written feedback in the LMS. Feedbacking in SERVE is taken at highest consideration to encourage students to undergo self-assessments thus allowing for selfimprovement [19]. The feedback is an authentic human interaction between the students and the teacher. From the student’s view, the feedback validated their strengths as writers.

3 Conclusion The online module using the SERVE model promoted students’ self-directed learning and delimited the students’ low internet connectivity and inadequate home learning environment. The concise instructor-made and open education videos aroused the student’s interest. The teacher’s feedback became alternative formative and summative evaluation which impacted the quality of student outputs. The study recommends that high schoolaged students with connectivity problems can effectively learn through a self-directed module that follows SERVE model. The study can benefit from a more quantitative means of gathering data over larger class sizes to check the effectiveness of the model in a more massive online learning environment.

References 1. Budrikis, Z.: Physics in a time of COVID-19. Nat. Rev. Phys. 2, 177 (2020). https://doi.org/ 10.1038/s42254-020-0166-8 2. UNESCO: 1.37 Billion Students Now Home as COVID-19 School Closures Expand, Ministers Scale up Multimedia Approaches to Ensure Learning Continuity. https://en.une sco.org/news/137-billion-students-now-home-covid-19-school-closures-expand-ministersscale-multimedia 3. Cynthia, B.: OVPAA Memorandum No. 2020-68 (2020). https://up.edu.ph/wp-content/upl oads/2020/07/20200702-OVPAA-Final-cbb-June-22-Memorandum-2020-68-2.pdf 4. Iglesias-Pradas, S., Hernández-García, Á., Chaparro-Peláez, J., Prieto, J.L.: Emergency remote teaching and students’ academic performance in higher education during the COVID19 pandemic: a case study. Comput. Hum. Behav. 119, 106713 (2021). https://doi.org/10. 1016/j.chb.2021.106713 5. Pham, H., Tran, Q.-N., La, G.-L., Doan, H.-M., Vu, T.-D.: Readiness for digital transformation of higher education in the Covid-19 context: the dataset of Vietnam’s students. Data Brief. 39, 107482 (2021). https://doi.org/10.1016/j.dib.2021.107482 6. Munastiwi, E., Puryono, S.: Unprepared management decreases education performance in kindergartens during Covid-19 pandemic. Heliyon. 7, e07138 (2021). https://doi.org/10.1016/ j.heliyon.2021.e07138 7. Barrot, J.S., Llenares, I.I., del Rosario, L.S.: Students’ online learning challenges during the pandemic and how they cope with them: the case of the Philippines. Educ. Inf. Technol. 26(6), 7321–7338 (2021). https://doi.org/10.1007/s10639-021-10589-x

SERVE as Instructional Design

51

8. Kornpitack, P., Sawmong, S.: Empirical analysis of factors influencing student satisfaction with online learning systems during the COVID-19 pandemic in Thailand. Heliyon. 8, e09183 (2022). https://doi.org/10.1016/j.heliyon.2022.e09183 9. Agyeiwaah, E., Badu Baiden, F., Gamor, E., Hsu, F.-C.: Determining the attributes that influence students’ online learning satisfaction during COVID-19 pandemic. J. Hosp. Leis. Sport Tour. Educ. 100364 (2021). https://doi.org/10.1016/j.jhlste.2021.100364 10. Mamun, M.A.A., Lawrie, G., Wright, T.: Instructional design of scaffolded online learning modules for self-directed and inquiry-based learning environments. Comput. Educ. 144, 103695 (2020). https://doi.org/10.1016/j.compedu.2019.103695 11. Al Mamun, M.A., Lawrie, G., Wright, T.: Exploration of learner-content interactions and learning approaches: the role of guided inquiry in the self-directed online environments. Comput. Educ. 178, 104398 (2022). https://doi.org/10.1016/j.compedu.2021.104398 12. Gumalal, J., Vilbar, A., Bernardez, F.: Exploring a flexible blended learning model in technology deficient classroom. In: Theory and Practice of Computation, pp. 77–85. CRC Press (2020) 13. Fisher, T., Denning, T., Higgins, C., Loveless, A.: Teachers’ knowing how to use technology: exploring a conceptual framework for purposeful learning activity. Curric. J. 23, 307–325 (2012). https://doi.org/10.1080/09585176.2012.703492 14. Wang, Y., Cao, Y., Gong, S., Wang, Z., Li, N., Ai, L.: Interaction and learning engagement in online learning: the mediating roles of online learning self-efficacy and academic emotions. Learn. Individ. Differ. 94, 102128 (2022). https://doi.org/10.1016/j.lindif.2022.102128 15. Kanetaki, Z., et al.: Grade prediction modeling in hybrid learning environments for sustainable engineering education. Sustainability. 14, 5205 (2022). https://doi.org/10.3390/su14095205 16. Yaniawati, P., Kariadinata, R., Sari, N.M., Pramiarsih, E.E., Mariani, M.: Integration of elearning for mathematics on resource- based learning: increasing mathematical creative thinking and self-confidence. Int. J. Emerg. Technol. Learn. IJET. 15, 60 (2020). https://doi.org/ 10.3991/ijet.v15i06.11915 17. Zhu, M.: Enhancing MOOC learners’ skills for self-directed learning. Distance Educ. 42, 441–460 (2021). https://doi.org/10.1080/01587919.2021.1956302 18. Brame, C.J.: Effective educational videos: principles and guidelines for maximizing student learning from video content. CBE—Life Sci. Educ. 15, es6 (2016). https://doi.org/10.1187/ cbe.16-03-0125 19. Yang, A.C.M., Chen, I.Y.L., Flanagan, B., Ogata, H.: How students’ self-assessment behavior affects their online learning performance. Comput. Educ. Artif. Intell. 3, 100058 (2022). https://doi.org/10.1016/j.caeai.2022.100058

Extended Technology Acceptance Models for Digital Learning: Review of External Factors Akrivi Krouska(B)

, Christos Troussas , and Cleo Sgouropoulou

Department of Informatics and Computer Engineering, University of West Attica, Ag. Spyridonos Street, 12243 Egaleo, Greece {akrouska,ctrouss,csgouro}@uniwa.gr

Abstract. The rapid development of information and communication technologies (ICT) has revolutionized teaching and learning strategies, leading to digital learning technologies. Digital learning is any type of learning that uses technology, such as Internet, social media, web technologies, multimedia, mobile devices, etc. Since digital learning has dominated in education nowadays, measuring the acceptance of this technology and predicting the behavioral intention to use it is of great importance. The most established model for this purpose is the Technology Acceptance Model (TAM), consisting of certain constant predictors extended by external variables. To identify the most commonly used external factors of TAM in digital learning, an analysis of 21 studies from 2015 onwards was conducted. These researches were classified into three fields of digital learning, namely e-learning, m-learning, augmented/virtual reality systems, in order to analyze the different variables used depending on the kind of system evaluated. The results show that self-efficacy, perceived enjoyment and system quality are significant predictors of user attitude used regardless of the digital learning technology evaluated, followed by subjective norm, system accessibility and facilitating conditions. Keywords: Augmented reality · E-learning · Extended TAM · Mobile learning · Technology adoption · User acceptance · Virtual reality

1 Introduction Information and communication technologies (ICT) are increasingly being used in education, overcoming the time and place limits of face-to-face instruction and introducing new practices of teaching and learning [1, 2]. Thus, digital learning is considered an educational approach that provides flexibility in participating and ease of access in learning process to a wide range of audiences [3, 4]. Digital learning technologies exploit technology, such as web, social media, multimedia, animations, mobile devices etc., to provide online learning environments, enabling instructors to deliver learning material, as well as promoting communication and collaboration between peers [5, 6]. However, the benefits of such technologies cannot be appreciated if the developed systems are not used by learners. In order to evaluate the user acceptance of new learning environments and investigate the factors affecting the adoption of such technological tools, a number of theoretical © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 52–63, 2023. https://doi.org/10.1007/978-3-031-17601-2_6

Extended Technology Acceptance Models for Digital Learning

53

models have been introduced. The most diffused approach is the Technology Acceptance Model (TAM), proposed by Davis [7]. This model explores the predictors of technology acceptance, from the perspective of users’ perceptions about innovations, social and contextual factors. The core variables of TAM is the perceived ease of use and perceived usefulness, which are measured after the use of the system for a while in order to predict the behavioral intention and actual usage. However, digital learning researchers have been extending TAM with appropriate external factors related to the nature of the systems developed for testing their impact on behavioral intention and making more reliable system utilization predictions [8–10]. Hence, there are a large number of different external variables and a high number of extended TAM models in digital learning adoption studies. Several reviews and meta-analysis have been conducted investigating the different theoretical frameworks for assessing technology acceptance in education [11, 12] or exploring the extended TAM used in certain field of online learning, e.g. e-learning or mobile learning (m-learning) [13, 14], or examining the factors influencing specific type of users, like faculty members, teachers or students [15, 16]. However, to the best of authors’ knowledge, currently there is hardly a review and comparison of the external variables incorporating into TAM for evaluating the user acceptance of the three major fields of digital learning technologies, namely e-learning, m-learning and augmented/virtual reality environments. Thus, the main purposes of this study are to explore the appropriate predictors used emerging from the different nature of digital learning technologies, to identify the most commonly used external factors that play an important role in the adoption of new technology and to help researchers by affording practical determinants to them extracting from the literature that definitely affects learners in using digital learning environments.

2 The Technology Acceptance Model (TAM) Many research models have been used by various researchers for evaluating the technology acceptance. Among these models, the Technology Acceptance Model (TAM) is the most used and reported model in the social sciences [17], introduced by Davis in 1989 [7]. TAM was adapted from the Theory of Reasoned Action (TRA), which is the theoretical framework of explaining user behavior toward IT systems. TAM specifies that the user attitude or feeling, either positive or negative, about the behavioral intention toward adopting a system is affected by perceived usefulness of the system and its perceived ease of use [7]. Moreover, the perceived ease of use influences the perceived usefulness; while, the perceived usefulness affects behavioral intention to use. On the other hand, the actual use of the system is affected by user behavior intention to use the system [7]. In TAM, external and context-dependent factors are proposed to investigate their effect on users’ two main determinants, perceived ease of use and perceived usefulness. The extended TAM not only predicts technology acceptance, but also provides further explanation about system usage and better guidance concerning system development. Figure 1 illustrates TAM extended by external variables.

54

A. Krouska et al.

Perceived Usefulness Attitude toward Using

External Variables

Behavioral Intention to Use

Actual Use of System

Perceived Ease of

Fig. 1. Technology acceptance model (Davis 1989)

TAM has been widely used to underpin educational technology acceptance. The reason why the majority of the researchers prefer this model is due to its high credibility and simplicity in applying it [18]. Moreover, the capability of expanding TAM offers more effective explanation of the acceptance of technology [19].

3 Research Methodology This study reviewed the existing literature in order to find recent studies on digital learning technologies that have evaluated user acceptance using an extended TAM. For this purpose, Scopus search engine was used, applying the keywords such as “extended TAM e-learning”, “extended TAM mobile learning”, “extended TAM augmented reality in education” and “extended TAM virtual reality in education”. Moreover, the results were limited to include studies from 2015 onwards on the subject area of computer science. As such, the numbers of papers returned by the searches were 47, 17 and 13 regarding the field of e-learning, m-learning and AR/VR, respectively. Afterwards, a data analysis was conducted, resulting in 21 studies that were selected to be included in the review, namely 7 studies per field, based on the following criteria: i. the papers should investigate student acceptance of technology adoption; thus, papers that explored only teacher perspectives were excluded, ii. the methodology used should be well described, iii. an empirical study should be conducted, and iv. the results should be clearly presented. After identifying the valid papers included in the review, the external factors used were listed and analyzed. Table 1 illustrates the 21 studies that were identified and included in current review. The majority of the papers were published in respected international journals (67%), while the rests in significant international conferences (33%). The mean of the number of external variables used extended TAM was 3, while their frequency was: 2 external variables used in 42.8% of the studies, 3 in 28.6%, 4 in 19%, 5 and 6 in 4.8%. Moreover, in 81% of the researches, the sample consisted of more than 200 participants, making the findings more reliable. Finally, regarding the country the studies took place, 4 of them were conducted in Greece and China, 2 in United Arab Emirates and Malaysia, 1 in India, Dubai, Yemen, Spain, Indonesia, Brunei Darussalam, South Africa, U.S. and Jordan.

Extended Technology Acceptance Models for Digital Learning

55

Table 1. Review of 21 studies using extended TAM to evaluate digital learning technology adoption. Study Title

Paper type

Year

Country

Users

# Field Ext. Var.

[10]

Impact of social networking Journal for advancing learners’ knowledge in E-learning environments

2021 Greece

200 undergraduate students

3

e-learning

[20]

Examining the structural relationships among e-learning interactivity, uncertainty avoidance, and perceived risks of COVID-19: Applying extended technology acceptance model

Journal

2022 India

288 undergraduate and graduate students

3

e-learning

[21]

Acceptance of E-learning among university students in UAE: A practical study

Journal

2020 United Arab 366 university Emirates students

3

e-learning

[22]

Understanding the Impact of Social Media Practices on E-Learning Systems Acceptance

Conference 2019 Dubai

410 graduate and undergraduate students

3

e-learning

[23]

Students’ intentions to use PBWorks: a factor-based PLS-SEM approach

Journal

2019 China

429 junior secondary students

2

e-learning

[24]

Investigating Users’ Journal Continued Usage Intentions of Online Learning Applications

2019 China

275 4 respondents included college students and staff

e-learning

[25]

Evaluating the early adoption of Moodle at a higher education institution

181 undergraduate students

e-learning

[26]

Attitude Towards Intention Conference 2021 Malaysia to Use Mobile-Based Teaching Assessment Based on TAM and a Novel Extended TAM Model

75 participants 2 of undergraduates, graduates studies, and academic support

m-learning

[27]

Factors Influencing Conference 2021 Yemen Students’ Intention To Use Mobile Learning: A Study at Yemen Higher Education Institutions

381 university students

2

m-learning

[28]

An extended technology acceptance model of a mobile learning technology

293 undergraduate students

3

m-learning

Conference 2015 South Africa

Journal

2019 East China

4

(continued)

56

A. Krouska et al. Table 1. (continued)

Study Title

Paper type

Year

Country

[29]

Understanding the Quality Journal Determinants that Influence the Intention to Use the Mobile Learning Platforms: A Practical Study

[30]

An Extended Technology Conference 2021 Spain Acceptance Model in the Context of Mobile Learning for Primary School Students

[31]

Users

2019 United Arab 221 university Emirates students

# Field Ext. Var. 4

m-learning

162 primary school students

2

m-learning

The Impact of Personality Conference 2020 Malaysia Traits Towards the Intention to Adopt Mobile Learning

351 university students

2

m-learning

[32]

A study on senior high Journal school students’ acceptance of mobile learning management system

2019 Indonesia

300 students in senior secondary education

5

m-learning

[9]

Exploring Users’ Behavioral Intention to Adopt Mobile Augmented Reality in Education through an Extended Technology Acceptance Model

Journal

2022 Greece

220 secondary school students

2

Augmented/ Virtual reality

[33]

Using virtual reality for dynamic learning: an extended technology acceptance model

Journal

2022 U.S.

489 flight students

6

AR/VR

[34]

Student Acceptance and Attitude Towards Using 3D Virtual Learning Spaces

Conference 2016 Brunei Darussalam

85 university students

2

AR/VR

[35]

User acceptance of augmented reality welding simulator in engineering training

Journal

2022 Greece

200 trainees

2

AR/VR

[36]

Virtual reality technology in Journal construction safety training: Extended technology acceptance model

2022 China

1158 construction engineering employees and students

3

AR/VR

[37]

Factors influencing behavior intentions to use virtual reality in education

Journal

2022 Jordan

503 university students and lecturers

4

AR/VR

[38]

Measuring user experience, usability and interactivity of a personalized mobile augmented reality training system

Journal

2021 Greece

200 trainees

2

AR/VR

Extended Technology Acceptance Models for Digital Learning

57

4 Results and Discussion This review paper revealed that the authors of the selected researches have extended TAM with different predictors depending on the scope of their study and the special features of the technology used. Regarding the external factors used in evaluating e-learning adoption, it is observed that the most common one is the facilitating conditions, namely the degree to which an individual believes that an organizational and technical infrastructure exists to support the use of the system. Providing technical support to users can lead to greater acceptance and to the successful adoption of the new technology. Other external variables met in several studies are social features incorporated into the system, gamification approaches used to engage students in learning process, subjective norm related to social influence to adopt the usage of the system, system quality regarding the information, content, response and good operation of the system, and individual differences related to user convenience of using technology. Table 2 illustrates all the external predictors recorded in e-learning studies. Table 2. External factors used by studies related to e-learning. External variable

Description

Study

Social network usage

The time people spend in social networking activities [10]

Social features

The social media interaction, the willingness to [10, 22] “follow” or “friend” other users, as well as to react on the posts/images of other users

Gamification

The application of game design-like elements and game principles in non-game contexts

[10, 22]

Interactivity

Interactive communication between peers within system

[20]

Uncertainty avoidance

The extent to which individuals in a society feel threatened by situations that are undefined, ambiguous, and uncertain

[20]

Perceived risks of COVID-19 User’s subjective evaluation related to their risk of an [20] illness or an adverse outcome often in relation to performing certain risky behavior Enjoyment

The activity of using a specific system is perceived to [21] be enjoyable

Accessibility

The degree of ease of how a user can access and use the information and extracted from the system

[21]

Subjective norm/social influence

The belief about whether most people approve or disapprove of the behavior

[21, 24]

Knowledge sharing

The business processes that distribute knowledge among all individuals participating in process activities

[22]

(continued)

58

A. Krouska et al. Table 2. (continued)

External variable

Description

Study

Technical support/facilitating The provision of technological equipment, facilities, conditions online services and tools

[23–25]

Teacher support

Support students’ use Of computers by providing them with computer skill training

[23]

System characteristics (Quality)

Job relevance, output quality, and result demonstrability

[24, 25]

Individual differences

Computer self-efficacy, computer anxiety, computer playfulness, and perception of external control

[24, 25]

Regarding m-learning technology, the most common external predictors extending TAM for the analysis of user acceptance are perceived enjoyment indicating how pleasant and interesting the students find the system, self-efficacy related to user confidence in using technology, subjective norm referring to the influence on using technology by the surrounding and system quality. In Table 3, all the external variable used in m-learning studies are listed. Table 3. External factors used by studies related to m-learning. External variable

Description

Study

System quality

Information/content/service quality and availability of system

[26, 27, 29]

Perceived enjoyment

Enjoyment while using the system

[26, 28, 30, 31]

Superior influence

Teacher support to use the system

[28]

Self-efficacy/Perceived convenience Individual’s self-confidence in utilizing technological innovation

[30–32]

Personal innovativeness

Willingness to adopt the technology

[32]

Subjective norm/Social influence

An individual’s attitudes, beliefs or [27, 28, 32] behavior are modified by the presence or action of others

Relative advantage

Degree to which using the current system is perceived to be superior to its predecessor

[32]

System accessibility

Free access and use of the system

[32]

Regarding augmented and virtual reality, the analysis of external variables of extended TAM models shows that the most common used predictors are perceived enjoyment, self-efficacy, system quality, perceived playfulness related to gamification

Extended Technology Acceptance Models for Digital Learning

59

characteristics embodied into the system and perceived behavioral control referring to users’ perceptions of their ability to perform a given behavior. Table 4 shows all the external predictors used in AR/VR studies. Table 4. External factors used by studies related to AR/VR. External variable

Description

Study

Quality output

Technological appropriateness of teaching with AR

[9, 35]

Perceived playfulness

Gamification characteristics

[9, 36]

Perceived enjoyment

Fun to use the technology

[33–35]

Perceived interactivity

User engagement with media content

[38]

Perceived personalization

Provide dynamic learning based on users’ needs

[38]

Performance expectancy

VR technology will improve performance as compared to another training device

[33]

Perceived health risk

Health concerns due to VR usage

[33]

Regulatory uncertainty

The lack of regulations regarding the use of VR

[33]

Self-efficacy

User’s individual judgment of how well [33, 34, 36] a course of action can be executed in a prospective situation

Perceived behavioral control/Perceived The perceptions formed about the ease Effort Expectancy or difficulty of using VR

[33, 37]

Perceived price value

The economic burden of implementing [36] a well-established system and related equipment

Usability

The quality of a user’s experience when interacting with system

[37]

Perceived compatibility

Degree to which a new technology meets the habits, values and needs of user

[37]

Perceived facilitating conditions

The degree to which a person believes [37] that the existing organizational and technical infrastructure can support the use of technology

Comparing the external factors used in the three fields of digital learning technologies, it is observed that self-efficacy, perceived enjoyment and system quality are significant predictors of user attitude used regardless of the digital learning technology evaluated, followed by subjective norm, system accessibility and facilitating conditions. Selfefficacy is an important factor that influences the technology adoption; and especially in

60

A. Krouska et al.

learning environments, it can positively affect students’ motivation, concentration and learning outcomes. Enjoyment plays also an important role in educational technology adoption, since providing an enjoyable learning environment may motivate students to get involved in learning process, having positive results in increasing their participation and consequently their performance. System quality and accessibility are other factors that can positively affect the adoption of new technology, since allow users to have access to the system constantly, navigate in it and use its resources efficiently. Subjective norm can potentially impact technology adoption, especially in learning environments where teachers, peers and family members may strongly affect the students’ intention of utilizing and accepting technology to learn. Finally, the importance of facilitating conditions is evident in the education sector. Providing technical support to students can positively affect their intention to use the new technology.

5 Conclusions In recent years, information and communication technologies (ICT) have played a significant role in all aspects of modern society. Especially, in education, the emergence of internet and web technologies has introduced new ways of delivering instructional material and supporting teaching and learning, leading to digital learning. Digital learning means learning using digital age technologies (computer, mobile devices, internet, etc.). Examples of digital learning technologies are e-learning, m-learning and augmented/virtual reality. However, the benefits of digital learning cannot be maximized if students do not use it. As such, it is of great importance to analyze the factors that affect learners to use digital learning systems for resulting in an effective educational environment. For this purpose, a number of different technology adoption theories have been established, among which the most widely used framework is the Technology Acceptance Model (TAM). Digital learning researchers constantly use TAM, extending with different external variables according to the nature of the technology used. Hence, a high number of extended technology acceptance models and a large number of different external factors have been introduced for evaluating user acceptance in digital learning technology adoption. The aim of this study is to compare the technology acceptance of three digital learning technologies, namely e-learning, m-learning and augmented/virtual reality, and analyze the external variables used. As such, an analysis of 21 studies, 7 studies per digital learning field, from 2015 onwards was conducted. The results show that self-efficacy, perceived enjoyment and system quality are significant predictors of user attitude used regardless of the digital learning technology evaluated, followed by subjective norm, system accessibility and facilitating conditions. Part of our future work is to further enrich the review with new studies and analyze the relationship between the external variables and constant determinants of TAM. Moreover, another interesting research area is to explore and propose some possibly significant predictors that still have not been adequately examined.

Extended Technology Acceptance Models for Digital Learning

61

References 1. Troussas, C., Krouska, A., Alepis, E., Virvou, M.: Intelligent and adaptive tutoring through a social network for higher education. New Rev. Hypermedia Multimedia 26(3–4), 138–167 (2020) 2. Troussas, C., Krouska, A., Virvou, M.: A multilayer inference engine for individualized tutoring model: adapting learning material and its granularity. Neural Comput. Appl. 1–15 (2021). https://doi.org/10.1007/s00521-021-05740-1 3. Troussas, C., Krouska, A., Sgouropoulou, C.: Collaboration and fuzzy-modeled personalization for mobile game-based learning in higher education. Comput. Educ. 144, 103698 (2020) 4. Troussas, C., Krouska, A., Sgouropoulou, C.: Enriching mobile learning software with interactive activities and motivational feedback for advancing users’ high-level cognitive skills. Computers 11(2), 18 (2022) 5. Krouska, A., Troussas, C., Voulodimos, A., Sgouropoulou, C.: A 2-tier fuzzy control system for grade adjustment based on students’ social interactions. Exp. Syst. Appl. 117503 (2022) 6. Krouska, A., Virvou, M.: An enhanced genetic algorithm for heterogeneous group formation based on multi-characteristics in social-networking-based learning. IEEE Trans. Learn. Technol. 13(3), 465–476 (2019) 7. Davis, F.D.: Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Q. 13(3), 319–340 (1989). https://doi.org/10.2307/249008 8. Abdullah, F., Ward, R.: Developing a General Extended Technology Acceptance Model for ELearning (GETAMEL) by analysing commonly used external factors. Comput. Hum. Behav. 56, 238–256 (2016). https://doi.org/10.1016/j.chb.2015.11.036 9. Papakostas, C., Troussas, C., Krouska, A., Sgouropoulou, C.: Exploring users’ behavioral intention to adopt mobile augmented reality in education through an extended technology acceptance model. Int. J. Hum. Comput. Interact. 1–9 (2022) 10. Troussas, C., Krouska, A., Sgouropoulou, C.: Impact of social networking for advancing learners’ knowledge in E-learning environments. Educ. Inf. Technol. 26(4), 4285–4305 (2021). https://doi.org/10.1007/s10639-021-10483-6 11. Imtiaz, M.A., Mirhashemi, M.T.: Analyzing trends in technology acceptance studies in education domain. In: International Conference on Current Trends in Information Technology (CTIT), pp. 23–27. IEEE (2013). https://doi.org/10.1109/CTIT.2013.6749472 12. Grani´c, A., Maranguni´c, N.: Technology acceptance model in educational context: a systematic literature review. Br. J. Edu. Technol. 50(5), 2572–2593 (2019). https://doi.org/10.1111/ bjet.12864 13. Gao, H.L.: A Systematic literature review of technology acceptance model and theory of planned behaviour towards online learning perspective. J. Arts Hum. 8(11), 75–82 (2019) 14. Al-Emran, M., Mezhuyev, V., Kamaludin, A.: Technology acceptance model in M-learning context: a systematic review. Comput. Educ. 125, 389–412 (2018) 15. Scherer, R., Siddiq, F., Tondeur, J.: The technology acceptance model (TAM): a meta-analytic structural equation modeling approach to explaining teachers’ adoption of digital technology in education. Comput. Educ. 128, 13–35 (2019). https://doi.org/10.1016/j.compedu.2018. 09.009 16. Falah, J., Alfalah, S.F., Alfalah, T., Qutechate, W., Ayyoub, H., Muhaidat, N.: An analysis of the technology acceptance model in understanding the university of Jordan’s students behavioral intention to use m-learning. Int. J. Psychosoc. Rehabil. 24(09), 1297–1312 (2020) 17. Teo T.: Technology acceptance research in education. In: Teo T. (ed.) Technology Acceptance in Education. SensePublishers (2011). https://doi.org/10.1007/978-94-6091-487-4_1

62

A. Krouska et al.

18. King, W.R., He, J.: A meta-analysis of the technology acceptance model. Inf. Manage. 43(6), 740–755 (2006) 19. Taherdoost, H.: A review of technology acceptance and adoption models and theories. Procedia Manuf. 22, 960–967 (2018). https://doi.org/10.1016/j.promfg.2018.03.137 20. Girish, V.G., Kim, M.Y., Sharma, I., Lee, C.K.: Examining the structural relationships among e-learning interactivity, uncertainty avoidance, and perceived risks of COVID-19: applying extended technology acceptance model. Int. J. Hum. Comput. Interact. 38(8), 742–752 (2022) 21. AlHamad, A.Q.M.: Acceptance of E-learning among university students in UAE: a practical study. Int. J. Electric. Comput. Eng. 10(4), 2088-8708 (2020) 22. Salloum, S.A., Al-Emran, M., Habes, M., Alghizzawi, M., Ghani, M.A., Shaalan, K.: Understanding the impact of social media practices on e-learning systems acceptance. In: Hassanien, A., Shaalan, K., Tolba, M. (eds.) International Conference on Advanced Intelligent Systems and Informatics. AISC, vol 1058, pp. 360–369. Springer, Cham (2019). https://doi.org/10. 1007/978-3-030-31129-2_33 23. Cheng, E.W., Chu, S.K., Ma, C.S.: Students’ intentions to use PBWorks: a factor-based PLS-SEM approach. Inf. Learn. Sci. (2019) 24. Ji, Z., Yang, Z., Liu, J., Yu, C.: Investigating users’ continued usage intentions of online learning applications. Information 10(6), 198 (2019) 25. Govender, I., Rootman-le Grange, I.: Evaluating the early adoption of Moodle at a higher education institution. In: European Conference on e-Learning, p. 230. Academic Conferences International Limited (2015) 26. Abdirahman, A.A., Jabar, M.A., Hashi, A.O., Elmi, M.A., Rodriguez, O.E.R.: Attitude towards intention to use mobile-based teaching assessment based on TAM and a novel extended TAM model. In: Saeed, F., Mohammed, F., Ghaleb, F. (eds.) IRICT 2021. LNDECT, vol. 127, pp. 583–593. Springer, Cham (2022). https://doi.org/10.1007/978-3-030-987411_48 27. Basurra, S., Bamansoor, S.: Factors influencing students’ intention to use mobile learning: a study at Yemen higher education institutions. In: 2021 2nd International Conference on Smart Computing and Electronic Enterprise (ICSCEE), pp. 206–211. IEEE (2021) 28. Zhonggen, Y., Xiaozhi, Y.: An extended technology acceptance model of a mobile learning technology. Comput. Appl. Eng. Educ. 27(3), 721–732 (2019) 29. Alshurideh, M., Salloum, S.A., Al Kurdi, B., Monem, A.A., Shaalan, K.: Understanding the quality determinants that influence the intention to use the mobile learning platforms: a practical study. Int. J. Interact. Mobile Technol. 13(11) (2019) 30. Eskandari, S., Valente, J.P.: An extended technology acceptance model in the context of mobile learning for primary school students. In: Auer, M.E., Tsiatsos, T. (eds.) IMCL 2021. LNNS, vol. 411, pp. 269–280. Springer, Cham (2022). https://doi.org/10.1007/978-3-03096296-8_25 31. Nabipour Sanjebad, N., Shrestha, A., Shahid, P.: The impact of personality traits towards the intention to adopt mobile learning. In: Sharma, S.K., Dwivedi, Y.K., Metri, B., Rana, N.P. (eds.) TDIT 2020. IAICT, vol. 618, pp. 182–193. Springer, Cham (2020). https://doi.org/10. 1007/978-3-030-64861-9_17 32. Stefanus, Mauritsius, T.: A study on senior high school students’ acceptance of mobile learning management system. J. Theor. Appl. Inform. Technol. 97(13), 3638–3649 (2019) 33. Fussell, S.G., Truong, D.: Using virtual reality for dynamic learning: an extended technology acceptance model. Virtual Reality 1–19 (2021). https://doi.org/10.1007/s10055-021-00554-x 34. Abdul Hamid, H.H., Ahmad Sherjawi, Z., Omar, S., Phon-Amnuaisuk, S.: Student acceptance and attitude towards using 3D virtual learning spaces. In: Phon-Amnuaisuk, S., Au, T.-W., Omar, S. (eds.) CIIS 2016. AISC, vol. 532, pp. 107–118. Springer, Cham (2017). https://doi. org/10.1007/978-3-319-48517-1_10

Extended Technology Acceptance Models for Digital Learning

63

35. Papakostas, C., Troussas, C., Krouska, A., Sgouropoulou, C.: User acceptance of augmented reality welding simulator in engineering training. Educ. Inf. Technol. 1–27 (2021). https:// doi.org/10.1007/s10639-020-10418-7 36. Zhang, M., Shu, L., Luo, X., Yuan, M., Zheng, X.: Virtual reality technology in construction safety training: extended technology acceptance model. Autom. Constr. 135, 104113 (2022) 37. AL-Oudat, M., Altamimi, A.: Factors influencing behavior intentions to use virtual reality in education. Int. J. Data Network Sci. 6(3), 733–742 (2022) 38. Papakostas, C., Troussas, C., Krouska, A., Sgouropoulou, C.: Measuring user experience, usability and interactivity of a personalized mobile augmented reality training system. Sensors 21(11), 3888 (2021)

Extended Reality and Games

Designing a VR Application for Typhoon Preparedness Training in a Classroom Barbara P. David, Neil Jherome L. Hernandez(B) , Kim Farhant S. Palao, Richelle Ann B. Juayong, and Jaime D. L. Caro University of the Philippines Diliman, 1101 Quezon City, Philippines [email protected]

Abstract. In the Disaster Readiness and Risk Reduction Senior High School course under the Philippine K-12 curriculum, there is little handson application of lessons learned for disaster preparedness. There are existing virtual reality applications that have been designed to train disaster preparedness. However, issues related to user experience and the missed potential of VR in hands-on training were found. Most were also not designed for the classroom. In this paper, a design for a VR application that would address said issues and be intended for Senior High School students and teachers is proposed. This is meant to supplement the classroom experience. Features such as the usage of virtual hands, movement options, and recording of the user’s activity for future viewing, among others, are incorporated. Keywords: Disaster preparedness games

1

· Immersive virtual reality · Serious

Introduction

The Philippines has been plagued by natural disasters since time immemorial due to its geographical location. From 2000–2019, it is estimated that 149 million Filipinos had been affected by such disasters [5]. The most common hazard faced by Filipinos is the tropical cyclone, or the typhoon. An average of 20 typhoons visit the Philippines each year, each leaving a trail of destruction and human despair in its wake [14]. As a result of these disasters, RA 10121 or the Philippine Disaster Risk Reduction and Management Act of 2010 was enacted. Along with the formation of the National Disaster Risk Reduction and Management Council (NDRRMC), the institutionalization of Disaster Readiness and Risk Reduction (DRRR) education was decreed by the Republic Act and further elevated by the Department of Education Order 37, series of 2015 [8]. In alignment with the enactment of the Philippine Disaster Risk Reduction and Management (DRRM) Act of 2010 or RA 10121, the Department of Education (DepEd) created the Comprehensive DRRM in Education Framework in 2015 in order to set their direction and priorities in their DRRM programs. Though such programs had already existed before, the lack of framework c The Author(s), under exclusive license to Springer Nature Switzerland AG 2023  A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 67–76, 2023. https://doi.org/10.1007/978-3-031-17601-2_7

68

B. P. David et al.

resulted in uncoordinated efforts that may not have been in line with the needs of the education sector, thus necessitating the creation of the said framework [6]. The current framework consists of three pillars (safe learning facilities, school disaster management, and DRR in education) involved with the four DRRM thematic areas defined in RA 10121 (Prevention and Mitigation, Preparedness, Response, and Recovery and Rehabilitation) which should reinforce the Department’s desired outcomes of accessibility, quality education, and good governance. One of the ways this department order manifests is in the integration of DRRM education in the new K-12 curriculum, which was first implemented in 2016. A course named the Disaster Readiness and Risk Reduction (DRRR) subject was added, and it may be taken by students of the Science, Technology, Engineering, and Mathematics (STEM) track [13]. In this subject, students should be able to learn about the basic scientific concepts behind hazards, risks, and disasters, as well as the different kinds of hazards and the measures that one should take in order to reduce or mitigate the harm that can be brought by them [7]. Intriguingly, a study has found a weak and negative relationship between science literacy and disaster preparedness [3]. Science literacy, in this study, refers to knowledge of natural disasters. According to them, the curricular materials on the topic focused more on the two lowest levels of Bloom’s Taxonomy, which are remembering and understanding. Bloom’s Taxonomy refers to a hierarchy of cognitive processes that increase in complexity. These processes are believed to facilitate learning. It was posited that behavior (disaster preparedness) is associated more with higher order thinking skills, which means that application, which is after remembering and understanding on the taxonomy, should be given more focus in order for such a curriculum to influence disaster preparedness behaviors. Perhaps because the formal institutionalization of DRRR education in the Philippines is still in its nascent stage, recent studies about its current state are few and far between. However, there are some smaller studies on disaster preparedness and related topics that were done in schools around the country. A study done at a selected Las Pi˜ nas High School, for example, showed that their SHS students report varying confidence in readiness and preparedness for disasters [12]. The students generally had some knowledge on disasters and reported that they were somewhat aware of and prepared for disasters to come. The researchers of the study attributed this to their knowledge gained in Earth and Life Science, which include various geologic and hydrometeorological processes and the reading and usage of hazard maps. The students, however, had lower scores in disaster risk perception, indicating that they did not perceive any major risk to happen to their home or locality. Based on the curriculum and existing teaching guides, the method of teaching in this subject seems to involve the discussion of a concept and then having the students reflect their knowledge of the topic through an activity or project [15]; the actual training or drills relating to disaster preparedness seem to be disconnected from the curriculum. Though students may apply their knowledge

VR App for Typhoon Preparedness Training

69

to the prompts given through these activities, this and the disconnect with the drills indicate a gap in the practical, hands-on application of lessons learned. Additionally, during disaster simulation drills, students do not get to make decisions on actions to take. They simply follow a script of the simulation which is predetermined for the scenario. This lessens the realism and immersion of the simulations since students don’t experience the full effects of their actions. Not all of the vast effects of a disaster and how the student’s decisions will affect the scenario are considered as there is no practical way to do so in the current method of content delivery [10]. Various research have been done aiming to resolve the gap in DRRR knowledge application of students through the use of Virtual Reality (VR) [2,4,9–11]. These all have their own strengths and weaknesses that will be taken into account for the current project. Typhoon Simulation VR, the project this research was inspired by, is split into two parts: collecting items for a survival kit and taking proper actions when outside during the typhoon [4]. It also contains adaptive features and personalization by asking the user information about their household before the game starts. Although the research showed that participants had improvement on the knowledge of the topic, the use of mobile VR with Google Cardboard still limited the player’s experience. There were also issues with the controls, as users found that the movement in-game was difficult and induced nausea. The Philippines Disaster Preparedness Simulator [10], another VR game set entirely in the Philippine context, starts with briefing the user on what to do before the main simulation starts. After the simulation, the player will be debriefed and be given feedback. This simulation includes three scenarios: earthquake in a school, flooding in a house, and typhoon in a barangay. In the virtual setting, the user is free to explore and interact with the environment. However, in the end, there is no system in place for behavioral analysis or to suggest what could have done better nor data on the effectiveness of the simulation. Other applications similar to the aforementioned VR games exist, but in summary, previous VR disaster preparedness simulations in the Philippines provide scenarios in the most common disasters experienced by the people living there including typhoons, floods, fire and earthquakes. The most common technologies in these simulations make use of either mobile devices with Google Cardboard for its cheapness and accessibility or a VR headset with a gaming laptop for its power. It appears that research in VR DRRM training have not as much explored the use of new standalone VR technology such as the Oculus Quest, which provide its own benefits such as a good balance between price and power [9]. The mainstreaming of VR applications in disaster preparedness may aid in addressing the gap in knowledge application as an opportunity for students to apply lessons in a simulated environment. However, issues were found in the existing VR applications found in research, namely that they do not seem to adequately address the issue of hands-on application, and that issues with accessibility or usability may hinder their potential as widespread tools for training.

70

B. P. David et al.

There were some reports of nausea or discomfort due to the movement or controls [4,11]. Others also appeared to not fully utilize the hands-on nature of VR. Most utilized the virtual environment to present a simulation of the hazard or disaster, but a number of them did not allow the user to explore the world freely; they were instead presented with multiple-choice decisions to make [2,9]. Most projects studied were also not designed specifically for a classroom setting, one exception designed with grade-school students in mind [10]. With these considerations, a design for a standalone VR system application for flood and storm preparedness training that is meant to be a supplementary tool in a Senior High School classroom is presented. It is aimed to be handson, accessible for first-time VR users, and useful for teachers in observing their students’ performance. For this paper, in Sect. 2, the objectives in designing this application are stated, followed by Sect. 3 and 4, which contain the theoretical framework and the design and plans for testing respectively. Finally, in Sect. 5, conclusions are presented.

2

Objectives of the Study

The goal of the study is to implement a hands-on standalone VR application for use in a SHS DRRR class. This research will focus on utilizing standalone VR since it has the power of tethered VR, while being more scalable. The VR application will be targeted to run on the Oculus Quest 2. Once this is done, the usability and the effectiveness of this application in conveying disaster preparedness concepts and training will be studied and analyzed. The specific aims of this project are to: 1. Develop a VR application having the following features: (a) Environments simulating different scenarios including: i. Preparation before the typhoon, with the intent to evacuate This scenario will teach the player about storm warnings, what items they should have in their go-bag, and various tasks to secure their house and belongings before the typhoon. This scenario takes places several days before the typhoon. ii. Staying at home during the typhoon This scenario starts with the typhoon already in the player’s area. With flood slowly getting higher and higher, the player must follow safety protocols in the house before the flood gets high enough. At the end, the player must find a safe place to rest. (b) Interactive and Immersive Environments Almost every object can be grabbed and held by the player’s hands using the Oculus Quest’s hand controllers. Drawers and cabinets can be opened and electrical switches can be flipped by the player. During the typhoon, the player will be able to see the heavy rain and hear as if they were in the middle of the typhoon.

VR App for Typhoon Preparedness Training

71

(c) Feedback system At the end of each scenario, a list of items will be shown about the player’s performance. A score will be given as a quantitative feedback with a remark to further give context to the score. Finally, some suggestions will be shown as to what the player could have done better for other future training and real life situations. This is designed to reinforce learning through application and from mistakes. (d) Support for Teachers The application will allow customization of the simulation for the teachers to better fit the lesson and the logistics of the subject. Data from the students’ playthroughs are collected and can be sent to the companion web application for the teachers. Here, the teacher can access guides, tips, and insights on the students’ data to make the application more accessible for teachers. 2. Conduct Tests to measure the effectiveness of the VR application (a) Heuristic Evaluations and Usability Testing During development, the design of the application will be iteratively tested. While going through he flow of the application, problems found are noted to fix along with any various tweaks to the experience. (b) Post-app Survey After users have finished playing the VR application, they will be asked to accomplish a survey based on the Phillips ROI Model of training evaluation This will include questions about the users’ reaction, learning, application and implementation of what they learned, perceived impact of the application to them. Afterwards, the Return on Investment will be measured to get the value of the application. The Phillips ROI Model is used as it looks on how the application can further be improved.

3

Theoretical and Conceptual Framework

According to a study, only about a third of Filipinos have undergone disaster preparedness in the past [1]. As for the DRRR curriculum, it mostly focuses on remembering and understanding the concepts around the subject. However, according to Bloom’s Taxonomy, that only covers the first two levels of the educational goals. Therefore, the researchers aim to develop an immersive VR application to train students in disaster preparedness (Fig. 1). This project will build on top of gameplay features that appear in Typhoon Simulation VR [4]. The resulting VR application of this project will train students with what to do before and during a disaster, both scenarios having their own stage. Before the disaster, students must make their go bag and ensure that their house is well prepared for the upcoming disaster. Afterwards, during the disaster, the students must now focus on making themselves safe amidst the rising flood.

72

B. P. David et al.

Fig. 1. The conceptual framework used as a basis for this project.

To test and evaluate the software, Heuristic Evaluation and Usability Testing will be employed. Heuristic Evaluation checks the design of the application, while Usability Testing checks how the user interacts with the application [16]. The simulation approximately lasts for 15 min. Each stage will have a 7.5-min timer to give a sense of urgency during the simulation and to ensure there is enough

VR App for Typhoon Preparedness Training

73

time for each student to undergo the simulation. On a theoretical environment of 30 students along with 5 available sets of VR equipment from the school, it would take up a total of 2 h for every student to take their turn including any orientation held before the training.

4

Application Design

The application starts on a main menu which contains the game options and the button to start the game. This menu is meant to be accessible only to the teachers for their set-up before the students use the application. When the start button is pressed, the application will now show the start screen and the device can now be handed to the students. The start screen waits for the student’s input to start the game before prompting them to input their name for grading and identification. Once the student has indicated their information, they will enter the virtual environment. The player is placed inside a conservatively sized house. The house includes a living room, kitchen, bedroom and bathroom. The player may also go out to the front yard of the house. The environment will have a different tone and sound effects for the two scenarios. The first scenario will have clear sunny skies outside while the second scenario will have dark clouds with heavy winds and rain as the typhoon has already reached the house. The player may move around the environment either by walking around the area physically or using in the in-game teleportation. Objects scattered around the house may be picked up by the player and placed in their backpack. Not all interactive objects are necessary to take or interact with and bringing unnecessary items will affect the player’s results. Aside from objects that can be picked up, other objects that can be interacted with include doors, windows, cabinets, power switches, plugs, and valves. The location of these objects inside the house will be randomized to avoid memorization and is ensured to spawn in a location that makes sense (e.g., no food or drinking water is located in the bathroom). The simulation starts at the first scenario or the pre-flood stage. During the first scenario, the player is tasked with taking proper preparations before the typhoon. No specific list of tasks will be shown for both the first and second scenario. Instead, it will be left to the player’s judgement on what they should do based on what objects they see in the virtual environment and what they have learned from the class. The task list guideline can be turned on by the teacher as one of the options in the main menu. One of the tasks consists of collecting various items for the emergency kit during the typhoon. These items include food, water, first aid kit, flashlight, batteries, matchsticks, medicine, important documents, etc. Other than building the emergency kit, the player must also clear the drainage outside of any blockage and trim down long and loose branches from the tree at the front yard. The first scenario ends when the timer runs out or when the player indicates that they are done. Afterwards, the player will go to the second scenario or the flood stage. Aside from heavy rains and dark clouds outside, there will be a flood slowly

74

B. P. David et al.

creeping up the house. This flood will rise slower if the player had cleared the drainage successfully during the first scenario. During the second scenario, the player’s tasks include closing all doors and windows, shutting off the main power switch, unplugging appliances, and turning off the gas tank. There are systems that detect if the player does dangerous actions like staying outside during the typhoon or swimming in the flood. Any dangerous actions detected will reflect on the feedback to the player’s performance. The scenario ends when the player indicates a safe place to rest for the rest of the typhoon or when the timer runs out. After the simulation, there will be a short briefing inside the program on how the player performed. Here, feedback is shown for each task that the player has done or has failed to do. The player will be warned on which task they have failed to do or have done insufficiently and be given positive feedback for the tasks done correctly. This will be followed by a final rating on the player’s performance calculated from the tasks they accomplished which can serve as their grade for the activity. Finally, the program goes back to the start screen and waits for the next student to start. In the main menu, the teacher can tweak various options for the simulation such as the list of tasks to be done, duration of each scenario, showing the specific task list, and exporting grades and data. When exporting the grades, the VR application will upload the data from the simulations to the PC application for the teachers. The data is in a table format displaying which tasks each student have done or not done (and quantity if applicable), possible dangerous actions the students took, and their total grade. There are also charts that will show the average quantity of food or water brought per student, average time spent for each task, and other statistics that might be of interest to the teacher. The teacher may use these data to conduct a post-activity discussion with their students on what they could improve on and what to avoid next time. A possible feature is to have a randomly generated house interior. Not only would this further reinforce learning by experience and not memorization, this would also make the application have high replayability and reusability for training. This will also make using the application as a form of a quiz more viable since the students’ knowledge on applying what they learned on any situation will also be tested. 4.1

Implementation

To develop the game, the following technologies will be used: – Oculus Quest 2 – Unity – Blender The Oculus Quest 2 is a VR headset that may or may not be used in conjunction with a desktop computer. In the case of this design, the application will utilize full standalone VR, which means that no connection to the desktop computer is required to run the application. The Oculus Quest 2 comes with two controllers which both will be required to use the application.

VR App for Typhoon Preparedness Training

4.2

75

Curriculum Integration

This application is intended to be a supplementary training for the class and would ideally have a lecture before and/or after the simulation. To integrate the training into the curriculum, the school has the choice of doing it either during class hours or on a dedicated separate time. Using the default list of tasks, it is estimated that each student will take 15 min to finish the simulation. This means that a school with 5 VR equipment sets and an average class size of 30 students will take up 1.5 h to finish the training sessions for each class. Of course, the teacher can change the list of tasks and the duration of the simulation to how they see fit. It is recommended that the task list guidelines is turned on for learning and turned off for training/quiz. 4.3

Effectiveness Assessment

To measure the effectiveness of the training, interviews as well as pre-tests and post-tests using the Phillips ROI Model. With this, it would be possible to see not only if the training is effective or not, but also to see why and how it can be further improved. This also tests for the possible return on investment for schools since they would have to invest in expensive VR sets.

5

Conclusion

This paper has discussed the objectives and the initial design and test plans for the development of a VR application that serves as a supplementary training tool for hydrometeorological hazard preparedness. It tries to create a scalable application with the DRRR classes of SHS students and teachers in mind by making it hands-on, accessible for first-time VR users, and useful for user performance monitoring.

References 1. Bollettino, V., Alcayna-Stevens, T., Sharma, M., Dy, P., Pham, P., Vinck, P.: Public perception of climate change and disaster preparedness: evidence from the Philippines. Clim. Risk Manag. 30, 100250 (2020). https://doi.org/10.1016/j. crm.2020.100250. https://www.sciencedirect.com/science/article/pii/S2212096320 300401 2. Caballero, A.R., Niguidula, J.D.: Disaster risk management and emergency preparedness: a case-driven training simulation using immersive virtual reality. In: Proceedings of the 4th International Conference on Human-Computer Interaction and User Experience in Indonesia, CHIuXiD 2018, pp. 31–37. Association for Computing Machinery, New York (2018). https://doi.org/10.1145/3205946.3205950 3. Cahapay, M.B., Ramirez, R.P.B.: Relationship between science literacy and disaster preparedness: the possible role of curriculum in behavior theories. Asian J. Sci. Educ. 2(2), 78–86 (2020). https://doi.org/10.24815/ajse.v2i2.16803

76

B. P. David et al.

4. Caro, J., Malinao, J., Valencia, A., Lopez, M.: Adaptive virtual reality disaster simulation for community training. In: ICERI2019 Proceedings of 12th Annual International Conference of Education, Research and Innovation, IATED, pp. 9071– 9079, November 2019. https://doi.org/10.21125/iceri.2019.2180 5. Centre for Research on the Epidemiology of Disasters, UN Office for Disaster Risk Reduction: Human cost of disasters. Technical report, Centre for Research on the Epidemiology of Disasters and UN Office for Disaster Risk Reduction, October 2020. https://philippines.un.org/en/95346-human-cost-disastersoverview-last-20-years-2000-2019. Accessed 14 Nov 2021 6. Department of Education: The comprehensive disaster risk reduction and management (DRRM) in education framework, DO 37, August 2015. https://www.deped. gov.ph/2015/08/12/do-37-s-2015-the-comprehensive-disaster-risk-reduction-andmanagement-drrm-in-basic-education-framework/ 7. Department of Education: Disaster readiness and risk reduction (2019). https:// www.deped.gov.ph/wp-content/uploads/2019/01/SHS-Core Disaster-Readinessand-Risk-Reduction-CG.pdf. Accessed 10 Dec 2021 8. Disaster Risk Reduction and Management Service: DRRMS strategic plan & achievements (2019). https://www.deped.gov.ph/wp-content/uploads/2020/11/ 09 TLM DRRMS-StratPlan 20190709.pdf. Accessed 14 Nov 2021 9. George, M., Oliva, E.: Immersive technologies & digital games for school disaster preparedness, August 2019 10. George, M., Oliva, E.: Philippines disaster preparedness simulator, August 2019 11. George, M., Oliva, E.: Virtual reality based disaster resilience training, August 2019 12. Mamon, M.A., Suba, R.A., Son Jr., I.: Disaster risk reduction knowledge of Grade 11 students: impact of senior high school disaster education in the Philippines. Int. J. Health Syst. Disaster Manag. 5(3), 69–74 (2017). https://doi.org/10. 4103/ijhsdm.ijhsdm 16 17. https://www.ijhsdm.org/article.asp?issn=2347-9019; year=2017;volume=5;issue=3;spage=69;epage=74;aulast=Catedral;t=6 13. Official Gazette: The k to 12 basic education program. https://www.officialgazette. gov.ph/k-12/. Accessed 10 Dec 2021 14. Philippine Atmospheric, Geophysical and Astronomical Services Administration: Tropical cyclone information (2021). https://www.pagasa.dost.gov.ph/climate/ tropical-cyclone-information. Accessed 14 Nov 2021 15. Philippine Normal University: Teaching guide for senior high school disaster readiness and risk reduction core subject (2016) 16. Tapia, E.: Heuristic evaluations and usability testing are critical to your business, October 2017. https://www.chaione.com/blog/heuristic-evaluationsusability-testing. Accessed 29 Nov 2021

Virtual Reality in Education: Reviewing Different Technological Approaches and Their Implementations Andreas Marougkas(B) , Christos Troussas, Akrivi Krouska, and Cleo Sgouropoulou Department of Informatics and Computer Engineering, University of West Attica, Egaleo, Greece {amarougkas,ctrouss,akrouska,csgouro}@uniwa.gr

Abstract. Virtual Reality is a beneficial and appealing supplement to traditional education. VR has been a broad term used to designate a variety of educational systems throughout the years, including desktop-based Virtual Reality, CAVE-based Virtual Reality, stereoscopic glasses-based Virtual Reality and custom-developed Virtual Reality systems. This study investigates the many applications of the term “Virtual Reality” for educational purposes, with the objective of serving as a map to differentiate earlier uses of the term “Virtual Reality” as it is currently interwoven with Head Mounted Displays. The study also discusses the features and advantages of modern Head Mounted Displays that provide a complete VR learning experience that impact the user’s experience. More degrees of freedom (DoF), increased display resolution, greater refresh rate, hand and head tracking and a larger field of vision (FoV) are just a few of the hardware advancements that are promoting the use of VR as a technological method for education in the classroom. Keywords: Virtual reality · Education · Head mounted displays

1 Introduction Virtual Reality [1], Augmented Reality [2], Extended Reality [3] and Mixed Reality [4] are examples of promising and emerging technology approaches to enhancing traditional learning methods inside the boundaries of the traditional classroom. Since its first appearance in the form that we know today, nearly four decades ago [5], VR has shown to be a promising technology. Despite the fact that virtual reality (VR) has only recently piqued the interest of users and researchers, it has a bright future ahead of it because various barriers, mostly technical, have been overcome. Major manufacturers like Meta, HTC, and Sony have invested in and focused on the present and future of VR technology with their most state-of-the-art products including the Oculus Quest 2, Vive Flow, and PlayStation VR2. In addition, Meta, through the Oculus App Store, Valve, through Steam, and Sony, through the PlayStation Store, help to this progress by constantly enhancing the content with software apps.

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 77–83, 2023. https://doi.org/10.1007/978-3-031-17601-2_8

78

A. Marougkas et al.

VR has already been used in a variety of fields such as marketing [6], engineering [7], health sciences [8], art [9], automotive [10], biology [11], chemistry [12], history [13] and many more. In the sphere of education, VR research is gaining popularity. VR is being used by educators to explore new ways of delivering learning content in conjunction with traditional teaching methods, with the objective of benefiting both learners and instructors. This study seeks to serve as a road map for distinguishing past applications of the word “Virtual Reality” from its contemporary connotation of Head Mounted Displays. A number of technologies known as “Virtual Reality” systems have been deployed in education during the previous decade, and this is the major focus of this study. The study also goes over the applied technologies as well as the characteristics of contemporary Head Mounted Displays, which enable a comprehensive Virtual Reality learning experience in the classroom.

2 A Diverse Range of VR Systems Over time, the term “virtual reality” has morphed into a concept that encompasses what are now known as Head Mounted Displays. Although, a variety of VR systems with various setups and characteristics have been used for educational purposes such as Desktop based VR, CAVE based VR, Stereoscopic glasses-based VR, Custom developed VR as described in this section. 2.1 Desktop Based Virtual Reality By combining a real-world setting with fully working artificial equipment that imitates real-world devices and offers measurements within the virtual simulation, Valdez et al. [14] established a virtual lab for undergraduate electrical engineering students that was designed and operated on a desktop computer. For university undergraduate students, Xu and Ke [15] constructed another virtual simulation in the context of math learning. The simulation was developed and designed by the authors and it incorporated gamified features such as intuitive game control, an interactive interface, a scenario-based sequence and incentive mechanisms to enhance the enjoyment factor. Makransky et al. [16] also developed a virtual simulation for students that run on a desktop computer and allowed them to engage in a real-life scenario of a virtual lab in which they can freely conduct all of the necessary tasks to finish it. Dialoguing with a virtual instructor, operating all of the lab’s virtual tools and finding knowledge via wikilinks related to the subject are all examples of simulation interactions. Students also had control over the simulation’s flow by deciding when and how to move on to the next action. It’s worth noting that this is the only study revealed that uses the term “Virtual Reality” and uses a desktop computer to project the virtual application outside of the time span of 2011 to 2016. During the course Geoinformatics Applications of The School of Spatial Planning and Development, Kaimaris et al. [17] built a virtual space representing the Faculty

Virtual Reality in Education: Reviewing Different Technological

79

of Engineering, Aristotle University of Thessaloniki (AUTH). The authors converted panoramic shots, acquired with a digital camera and processed in Adobe Photoshop, into 360° spherical images that were projected in a common web browser using image stitching software. Regardless of the fact that each of the aforementioned studies used the term “virtual reality,” none of them employed modern immersive virtual reality headsets. Instead, desktop computers with conventional screens with limited field of view and low immersion were operated with conventional mice and keyboards. 2.2 CAVE Based Virtual Reality Stratos et al. [18] used CAVE technology to combine game-based learning and virtual reality to investigate student awareness and interest in manufacturing processes as a predictor of a possible manufacturing engineering profession. Cruz-Neira et al. [19] created CAVE VR at the University of Illinois in 1992. CAVE VR requires a physical facility with three or six projection panels as well as 3D stereoscopic eyewear. CAVE VR is a great method to create an immersive environment. However, one significant disadvantage of this technical method is that it only allows for modest embodiment because users cannot see their hands represented in the digital domain as they can in VR with modern HMDs. This implies that when utilizing the CAVE’s unintuitive controllers, users may observe them in a real-world environment. Furthermore, the tracking system utilized by CAVE VR allows just one person to participate in the virtual experience at a time, preventing the development of applications that allow for simultaneous user collaboration and participation. Another issue with CAVE VR is the installation complexity, as it requires a range of equipment such as motion capture systems, projection surfaces and a physical area that has been expressly constructed to meet its standards, making it difficult and costly to install elsewhere. 2.3 Stereoscopic Glasses Based Virtual Reality Abulrub et al. [20] described a “Fully Immersive” 3D environment as one in which the head and hands interact and navigate through the virtual environment using an infrared camera real-time tracking system. Through two projectors that alternated at a high frequency rate, the system projected images onto the user’s polarized glasses. The device enabled stereoscopic viewing of the virtual elements, but it did not provide the full field of vision with multiple degrees of freedom (3-DoF or 6-DoF) as modern HMDs do, considering the term “fully immersive” redundant. Barrett and Hegarty [21] approached college students to conduct two experiments focusing on the dimensionality of the display device and the manipulation of the handheld interaction device for a study involving molecular chemistry and students’ spatial ability. Students used a custom-built desktop computer setup that included a display monitor and Nvidia 3D Vision Wireless Glasses that provided stereoscopic visualizations via a WorldViz Vizard software-generated mirror. The handheld device that was employed as a controller system, was a cylinder-shaped manipulation device that allowed pupils to engage with virtual items like molecular structures. The students were able to view and interact with the objects thanks to a desktop computer setup that generated

80

A. Marougkas et al.

stereoscopic visuals projected on a display monitor and captured by Nvidia 3D Vision Wireless Glasses. The setup was referred to as a “Desktop Virtual Reality system” by the authors, who used a definition of virtual reality that differed from current ones that use Head Mounted Devices and VR controllers in terms of development and apparatus. It’s worth noting that Nvidia ceased supporting 3D Vision’s drivers in 2019, resulting in the product’s demise. 2.4 Custom Developed Virtual Reality Systems Häfner et al. [22] demonstrated how Virtual Reality was implemented in an interdisciplinary project-based course for engineering graduate and undergraduate university students to improve their engineering abilities and learn new skills connected to Virtual Reality software and hardware. Students were divided into teams that researched, planned and developed VR prototype projects after attending required Virtual Reality courses such as lectures, demonstrations and lab exercises. The first project presented was a two-prototype driving simulator named “DRIVE”. The first prototype featured a powerwall display with an Advanced Realtime Tracking System for head tracking, as well as a commercial video game racing wheel set and 3DVIA Virtools software platform for development. In the second project students used reversed engineering methods in order to modify a real car as a control unit and they replaced 3DVIA Virtools software platform in their projects after they decided to develop their own platform “Poly VR” programming in C++ language combined with Matlab/Simulink. Microsoft Kinect was used as a head tracking hardware and a powerwall monitor was used for projection. The third project was a car cockpit configuration called “IC3” which was created with CAD software and included a CAVE VR version with three surfaces as a projection system, an Advanced Realtime Tracking system for head tracking and a flystick2 controller.

3 Contemporary Virtual Reality Head Mounted Displays The term “Virtual Reality” has been applied to a variety of systems over the years, as stated in the preceding section. AR applications [23], for example, mostly employ a smartphone, while head-mounted displays, similar to those seen in past decades since its debut in the 1960s with Sutherland’s “Sword of Damocles” [24], have become associated with current VR technology and the name “Virtual Reality”. In comparison to the systems presented in the preceding section, modern HMDs have a plethora of features that make it easier for the user to have more natural and intuitive experiences. All contemporary HMDs feature 6 degrees of freedom (6DoF), which allows users to interact with the artificial environment in a more intuitive way [25]. HMD displays include LCD, OLED, and AMOLED displays that reproduce a pleasing and adequate color spectrum. Furthermore, resolution, pixel density, refresh rates, and field of view (FoV) are the primary parameters that transmit visual material to the user, lowering the risk of visual fatigue and oculomotor symptoms such as cybersickness [26, 27] while also providing an immersive and realistic experience.

Virtual Reality in Education: Reviewing Different Technological

81

Moreover, head tracking and eye tracking technology [28] are two new technologies that let users navigate inside the virtual environment while simulating real-life functions and delivering a more natural and realistic sense of movement and vision. Hand tracking technology also provides a more realistic feeling of human-machine interaction by allowing users to interact with artificial objects inside an artificial environment without the need of controls, much as they would in the real world [29] The hardware specifications of modern HMDs are shown in Table 1. Table 1. Modern HMDs specifications. System

HTC Vive Pro 2 Oculus Quest 2

HP Reverb G2

PlayStation VR

Display Resolution (per eye) Refresh rate

LCD 2448x2448

LCD 1832x1920

LCD 2160x2160

OLED 960x1080

Up to 120 Hz

Up to 120 Hz

90 Hz

Up to 120 Hz

Hand Tracking

No

Yes

No

No

Standalone use Field of view

No

Yes

No

No

120°

89°

114º

100°

4 Conclusion New means of distributing educational content are being created as a result of technological advances [30]. Virtual Reality is one of these advancements, and as a technology approach to education, it allows the user to engage with the virtual environment, resulting in an optimal learning experience that is derived from traditional learning methods and acts more as a complement. This study discusses the several Virtual Reality systems that have been used in educational applications, as well as the characteristics and specs of contemporary Head Mounted Displays.

References 1. Marougkas, A., Troussas, C., Krouska, A., Sgouropoulou C.: A Framework for Personalized Fully Immersive Virtual Reality Learning Environments with Gamified Design in Education. Frontiers in Artificial Intelligence and Applications EbookVolume 338: Novelties in Intelligent Digital Systems, pp. 95–104 (2021). https://doi.org/10.3233/FAIA210080C 2. Papakostas, C., Troussas, C., Krouska, A., Sgouropoulou, C.: User acceptance of augmented reality welding simulator in engineering training. Educ. Inf. Technol. 1–27 (2021). https:// doi.org/10.1007/s10639-020-10418-7 3. McGuirt, J.T., et al.: Extended reality technologies in nutrition education and behavior: comprehensive scoping review and future directions. Nutrients 12, 2899 (2020). https://doi.org/ 10.3390/nu12092899

82

A. Marougkas et al.

4. Birt, J., Stromberga, Z., Cowling, M., Moro, C.: Mobile mixed reality for experiential learning and simulation in medical and health sciences education. Information 9, 31 (2018). https:// doi.org/10.3390/info9020031 5. Berkman, M.I.: History of virtual reality. In: Lee, N. (eds.) Encyclopedia of Computer Graphics and Games. Springer, Cham (2018). https://doi.org/10.1007/978-3-319-08234-9_1 69-1 6. Anderson, K., Laverie, D.: In the consumers’ eye: a mixed-method approach to understanding how VR-Content influences unbranded product quality perceptions. J. Retail. Consum. Serv. 67, 102977 (2022). https://doi.org/10.1016/j.jretconser.2022.102977 7. Barbara, S., Vincent, H., Karim, B., David, B.: BIM data flow architecture with AR/VR technologies: use cases in architecture, engineering and construction. Autom. Construct. 134, 104054 (2022). ISSN 0926-5805. https://doi.org/10.1016/j.autcon.2021.104054 8. Chiang, D.H., et al.: Immersive virtual reality (VR) training increases the self-efficacy of inhospital healthcare providers and patient families regarding tracheostomy-related knowledge and care skills. Medicine 2 101–28570. https://doi.org/10.1097/MD.0000000000028570 9. Hutson, J., Olsen, T.: Digital humanities and virtual reality: a review of theories and best practices for art history. Int. J. Technol. Educ. 4, 491–500 (2021). https://doi.org/10.46328/ ijte.150 10. Hernández-Chávez, M., et al.: Development of virtual reality automotive lab for training in engineering students. Sustainability 13, 9776 (2021). https://doi.org/10.3390/su13179776 11. García-Bonete, M.J., Jensen, M., Katona, G.: A practical guide to developing virtual and augmented reality exercises for teaching structural biology. Biochem. Molec. Biol. Educ. 47 (2018). https://doi.org/10.1002/bmb.21188 12. Anastasia, R., Alexey, K., Andrey, M., Artem, S., Nadezhda, M., Mikhail, K.: Orbital battleship: a multiplayer guessing game in immersive virtual reality. J. Chem. Educ. 97(11), 4184–4418 (2020). https://doi.org/10.1021/acs.jchemed.0c00866 13. James, C., Rhodora, A.: Impact of immersing university and high school students in educational linear narratives using virtual reality technology. Comput. Educ. 159(104005), 0360–1315 (2020). https://doi.org/10.1016/j.compedu.2020.104005 14. Valdez, M., Machado Ferreira, C., Martins, M., Maciel-Barbosa, F.: 3D Virtual Reality Experiments to Promote Electrical Engineering Education (2015). https://doi.org/10.1109/ITHET. 2015.7217957 15. Xu, X., Ke, F.: Designing a virtual-reality-based, gamelike math learning environment. Am. J. Distance Educ. 30, 27–38 (2016). https://doi.org/10.1080/08923647.2016.1119621 16. Makransky, G., Mayer, R., Veitch, N., Hood, M., Christensen, K., Gadegaard, H.: Equivalence of using a desktop virtual reality science simulation at home and in class. PLOS ONE 14, e0214944 (2019). https://doi.org/10.1371/journal.pone.0214944 17. Kaimaris, D., Stylianidis, E., Karanikolas, N.: Virtual Reality: developing a VR space for academic activities. ISPRS Ann. Photogram. Remote Sens. Spatial Inform. Sci. II-5, 197–201 (2014). https://doi.org/10.5194/isprsannals-II-5-197-2014 18. Stratos, A., Rentzos, L., Mavrikios, D., Georgoulias, K., Mourtzis, D., Chryssolouris, G.: A virtual reality application to attract young talents to manufacturing. Procedia CIRP 57, 134–139 (2016). https://doi.org/10.1016/j.procir.2016.11.024 19. Cruz-Neira, C., Sandin, D.J., DeFanti, T.A., Kenyon, R.V., Hart, J.C., The, C.A.V.E.: Audio visual experience automatic virtual environment. Commun. ACM 35(6), 64–72 (1992). https:// doi.org/10.1145/129888.129892 20. Abulrub, A.H., Attridge, A., Williams, Z.: Virtual reality in engineering education: the future of creative learning. Int. J. Emerg. Technol. Learn. 6 (2011). https://doi.org/10.3991/ijet.v6i4. 1766

Virtual Reality in Education: Reviewing Different Technological

83

21. Barrett, T., Hegarty, M.: Effects of interface and spatial ability on manipulation of virtual models in a STEM domain. Comput. Hum. Behav. 65 (2016). https://doi.org/10.1016/j.chb. 2016.06.026 22. Häfner, P., Häfner, V., Ovtcharova, J.: Teaching methodology for virtual reality practical course in engineering education. Procedia Comput. Sci. 25, 251–260 (2013). https://doi.org/ 10.1016/j.procs.2013.11.031 23. Papakostas, C., Troussas, C., Krouska, Sgouropoulou, C.: Exploration of augmented reality in spatial abilities training: a systematic literature review for the last decade. Inform. Educ. 20(1), 107–130 (2021). https://doi.org/10.15388/infedu.2021.06 24. Sutherland, I.E.: A head-mounted three dimensional display. In: Proceedings of the December 9–11, 1968, Fall Joint Computer Conference, Part I. AFIPS 1968 (Fall, part I), pp. 757–764. ACM, (cit. on p. 6) San Francisco, California (1968) 25. Chakareski, J., Khan, M.: Wifi-VLC dual connectivity streaming system for 6DOF multiuser virtual reality. In: Proceedings of the 31st ACM Workshop on Network and Operating Systems Support for Digital Audio and Video (NOSSDAV 2021), pp. 106–113. Association for Computing Machinery, New York, NY, USA (2021). https://doi.org/10.1145/3458306. 3460999 26. Caserman, P., Garcia-Agundez, A., Gámez Zerban, A., Göbel, S.: Cybersickness in currentgeneration virtual reality head-mounted displays: systematic review and outlook. Virtual Reality 25(4), 1153–1170 (2021). https://doi.org/10.1007/s10055-021-00513-6 27. Ramaser, C.A.N., El Jamiy, F., Reza, H.: A systematic survey on cybersickness in virtual environments. Computers 11(4), 51 (2022). https://doi.org/10.3390/computers11040051 28. Callahan-Flintoft, C., Barentine, C., Touryan, J., Ries, A.J.: A case for studying naturalistic eye and head movements in virtual environments. Front. Psychol. 12, 650693 (2021). https:// doi.org/ https://doi.org/10.3389/fpsyg.2021.650693 29. Khundam, C., Vorachart, V., Preeyawongsakul, P., Hosap, W., Noël, F.: A comparative study of interaction time and usability of using controllers and hand tracking in virtual reality training. Informatics 8(3), 60 (2021). https://doi.org/10.3390/informatics8030060 30. Papakostas, C., Troussas, C., Krouska, A., Sgouropoulou, C.: Measuring user experience, usability and interactivity of a personalized mobile augmented reality training system. Sensors 21(11), 3888 (2021). https://doi.org/10.3390/s21113888

Ready to Play - A Comparison of Four Educational Maze Games Elena Paunova-Hubenova1 , Yavor Dankov2(B) , Valentina Terzieva1 , Dessislava Vassileva2 , Boyan Bontchev2 , and Albena Antonova2 1 Institute of Information and Communication Technologies, Bulgarian Academy of Sciences,

Acad. G. Bonchev Street, Bl. 2, 1113 Sofia, Bulgaria {elena.paunova,valentina.terzieva}@iict.bas.bg 2 Faculty of Mathematics and Informatics, Sofia University “St. Kl. Ohridski”, 1164 Sofia, Bulgaria {yavor.dankov,ddessy,bbontchev,a_antonova}@fmi.uni-sofia.bg Abstract. The increasing use of game-based learning necessitates software platforms and technological tools to create serious video learning games. This article presents the APOGEE process for the automated creation of maze video learning games and describes the development methodology. This methodology has been applied to the design of four educational video maze games enriched with various mini-games. They are in different thematic areas – two considering Bulgarian history in the 12–13 and 18–19 centuries, one game presenting the national characteristics of the old Bulgarian carpet industry, and another game dedicated to the monumental cultural heritage preservation in the era of intensified climate changes. The paper briefly describes the four video games and also presents their creation process, with a discussion about the game design. Keywords: Educational video games · Maze · Puzzle · Game-based learning

1 Introduction Last decade, educational video games appeared to be a promising tool for achieving better learning engagement among digital natives, thanks to their interactivity and visual attractiveness. Much research outlines various benefits of the use of serious games, and especially video games, in educational settings [1, 2]. Pedagogical practice asserts that blending traditional and innovative approaches to teaching is the most efficient. Contemporary educational games can inspire students’ curiosity, provoke their interest and stimulate their motivation through competitive elements [3, 4]. Further, they can present learning content and didactic tasks appealingly and entertainingly. Also, they enable alternative and more involving ways of learning content presentation, thus causing the effect of “imperceptible learning”. Different video games have educational potential, as they can offer autonomous learning and learning by doing while exploring the game context [5]. Much research shows that educational games can be more efficient than traditional learning approaches. The advantages are related to the motivation and engagement of students and the resulting learning outcomes [1, 2, 6]. Enriched educational video maze games are specific maze games with several halls, traversal paths, © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 84–94, 2023. https://doi.org/10.1007/978-3-031-17601-2_9

Ready to Play - A Comparison of Four Educational Maze Games

85

and various built-in mini-games. The game-based learning (GBL) approach is beneficial for both knowledge acquisition and the assessment of students’ acquired knowledge [7]. Several studies explore application areas of serious games – education, healthcare, military, economics, social activities, etc. [1, 6]. Despite the many benefits of educational games, their wide usage is hampered by the lack of free games in various subject areas. In addition, most of the available software tools and platforms for designing and creating educational games require significant effort and solid knowledge and skills in information technology. Thus, this issue usually hinders teachers develop even simple computer games intended for a particular subject and learning goals. This issue can be solved using generators of different kinds of 3D video mazes with embedded learning tasks and learning content [8]. The applied approach’s novelty consists of formal textual descriptions of educational maze games enriched with various mini-games and their automatic generation. This paper explains how the method of game generation is applied to creating four educational video maze games enriched with mini-puzzles. It is structured as follows: Sect. 2 introduces survey findings on teachers’ views on educational maze games and describes the process and methodology for developing such games within the APOGEE project software platform. Next, descriptions of all mini-games integrated into maze video games are given. The paper briefly describes the designed four different thematic areas of video games and compares their characteristics.

2 Research Background The authors conducted an online survey among school and university teachers (N = 198) to explore their views on educational maze games, preferred types of mini-games to be built-in, and usability of the APOGEE platform for game creation [9, 10]. The questionnaire applies a 5-point Likert scale. Before the survey, the respondents must watch a short video presenting the sample educational game generated by the APOGEE platform. Then they were asked to play the same maze game to become acquainted with the built-in types of mini-games. Here are presented some unpublished survey findings. 2.1 Teachers Need Educational Video Games Teachers’ opinions for the age group of learners for which the educational video mazes games are appropriate are presented in Fig. 1, applying a scale from 1 to 5.

Fig. 1. Appropriateness of video maze games for training.

86

E. Paunova-Hubenova et al.

Based on their experience, the respondents consider these games most suitable for low secondary school students (ages 11–14 years), followed by primary and high secondary students. These findings correspond almost with the usual target group of educational games. Hence, such games are highly appropriate for GBL in schools and less applicable to other students. However, very few are available educational games focused on continuous education.

Fig. 2. The preferences of teachers to different type of mini games.

Respondents have to point out what types of mini-games they prefer mostly to be built-in in the educational maze games. Figure 2 reveals that they favour most questions for unlocking maze doors and collecting and grouping the found artefacts, followed by a quiz for knowledge evaluation, asking intelligent non-playable character (NPC) to draw knowledge from the Web, and a 2D puzzle with learning image. Thus, these minigames are perceived as appropriate for pedagogical purposes, presenting didactic tasks engagingly. However, the range of preferences is relatively small (0.46), so teachers can use all types of puzzle games for content introducing and testing depending on particular learning goals and context. The survey findings show a positive attitude of teachers toward GBL and their need for software platforms to create educational video games effortlessly (Fig. 3).

Fig. 3. Teachers’ opinions about the process of creation of educational video maze games.

Ready to Play - A Comparison of Four Educational Maze Games

87

The respondents also find appropriate all the mini-games considered for building into the maze game. In general, they are optimistic regarding the functionalities offered by the APOGEE platform for automatized generation of educational video games. The expressed teachers’ opinions give us the stimulus for enhancing the game construction platform. Thus, till now, we have created four prototypes of video games on different topics using the APOGEE platform. This paper presents these educational games. 2.2 The Process for Automatized Creation of Educational Maze Games Unlike other existing approaches for the online generation of educational games, like quizzes [11], the APOGEE education software platform provides the opportunity to quickly and easily design, create and automatically generate educational video maze games [12]. It is done with the help of specially designed software instruments, which assist and support game designers in creating and evaluating the designed video games within the APOGEE platform [13]. Figure 4 illustrates the creation process of the APOGEE educational video maze game. In general, the process of creating an educational video maze game in the platform consists of 9 steps the designer goes through.

Fig. 4. APOGEE process of creation of educational video maze games.

The game designer can use these tools (assistive and analytics) to take advantage of them and facilitate their work. As a result, these instruments contribute to the creation process of educational video maze games that are specially designed to meet all available requirements, such as standard and conceptual requirements [15] related to the exposure environment, its aspects and virtual design [16, 17], as well as requirements as an educational purpose and goals, learning user groups, game and didactic content, user gaming, and learning experience and many more. As a result, a created educational game is available and ready to play by the users for whom it is intended.

88

E. Paunova-Hubenova et al.

2.3 The APOGEE Methodology for Creation of Educational Maze Games The APOGEE Methodology for designing and developing educational video games aims to facilitate teachers and educational professionals to build 3D educational mazes enriched by 2D and 3D mini-games of various types [14] representing multiple puzzles and didactic tasks. The APOGEE explorative video game scenarios can be arranged in an imaginary world in several consecutive maze halls (similar to the popular escape room games). To achieve the mission goals, the player must pass through all maze halls, complete the mandatory and optional mini-games, and collect the maximum number of points (i.e., achieve the maximal result). The process of learning and playing should take a minimum time with maximum efficiency. To enter a new hall, the player must unlock the door by answering specific questions. The didactic content can be placed on the learning boards in the halls, in puzzle games, and hidden objects. After completing the maze, the player may opt between continuing the play, repeating the game, or finishing the play. The game development platform APOGEE [18] is based on a formal XML description of the enriched maze created in an incremental and iterative process. It presents the structure and interior of the maze, together with the learning multimedia content and presentation and settings of the didactic mini-games. The XML description, together with game graphics and audio resources, are supplied to a custom plugin installed in the Unity 3D editor [19] and responsible for the generation of 3D mazes. Thus, the design of APOGEE 3D educational video mazes follows these consecutive steps: 1. 2. 3. 4. 5. 6. 7. 8. 9.

Setting learning objectives of the game and explorative maze scenario; Defining learning content, formatting text, images, audio, and multimedia files; Design of the audiovisual layout of the maze halls; Choice of suitable didactic mini-games, built-in content, and allocating them in the maze; Building an XML document describing the maze and its mini-games; Generation of the maze in the Unity 3D editor; Fine-tuning the maze game in the Unity 3D editor (if needed); Building executable versions of the game and distributing them among the players; Testing and validation of the game prototypes with target users.

The following versions of the generated maze game can add personalization of the content and adaptation of task difficulty [20, 21]. Several online versions of the maze game can be built and deployed on the APOGEE Web server.

3 Description of Four Educational Maze Games The APOGEE platform educational maze games generated contain a set of mini-games in their halls. These mini-games present the learning tasks students need to solve to reach the next room and gain the maximum score. The knowledge necessary for performing the tasks can be found on the information panel placed on the hall walls. Here are presented the mini-games used in the four maze games: • In the “Open Sesame!” game, the players try to answer a question to unlock the door to the next hall. They type the chosen answer or its number in a textbox.

Ready to Play - A Comparison of Four Educational Maze Games

89

• The players fill in a Quiz with several questions to acquire game points for the final score or reach a trigger score for unlocking a door. • To play the “I see you” mini-game, the user must detect and click on trans-lucent objects thematically related to the learning matter in the maze halls. • Within the “Find me!” game, in the maze halls, the user must uncover many objects hidden in or behind larger ones. • In the “Memory” puzzle, players search for matching couples of cards placed in a grid with the blank side up. • Within the “Word soup” puzzle game, a set of letters is arranged in a gridiron. The player looks for words that are relevant to the subject matter and are arranged vertically, horizontally, or diagonally. • The player must organize a set of objects by providing features in the “Divide & Conquer” mini-game. The game objects used are thematically related to the educational domain and can be gained in other puzzles in the maze halls. • The user must arrange a two-dimensional instructional image by moving their elements to the proper locations in the “2D puzzle” game. • The FPS (First Person Shooter) mini-game requires the player to shoot at unanimated moving objects, for example, flying balloons. • In the “Roll a Ball” mini-game, the players move several balls (labelled with text or images) by rolling them to the matching places or objects on the floor. In addition to the described mini-games, some maze games presented below include NPC [22], who interacts with the player. The NPC gives information that can be necessary for completing the didactic tasks or curious facts related to the learning domain in the hall. He searches for educational content in a database or reliable websites. Next, this section presents four didactic maze games created by the APOGEE platform with their thematic domains, target groups of players, and screenshots (Fig. 5).

Fig. 5. Screenshots of the fourth educational maze games: a) “The carpet industry in Bulgaria”, b) “The Dynasty of Asenetsi”, c) “The Legacy of Valchan Voyvoda”, d) “Let Us Save Venice”.

90

E. Paunova-Hubenova et al.

The first described didactic maze game is the “Carpet Industry in Bulgaria” [23], dedicated to cultural-historical heritage preservation and Bulgarian ethnography. It presents the handicraft of carpets after the XVII century in three Bulgarian towns. The six halls are thematically related as follows: Introduction Hall, the rooms West and North West from the first loop are dedicated to carpet weaving in Chiprovtsi, the joint North room present the carpet industry in Sliven, and the last two halls (East and North East) – Kotel’s carpets. Five halls include hidden objects and “Roll a Ball” mini-games. The exception is the central hall which is a part of both loops. A recorded video with sample gameplay is available at https://youtu.be/ZLH4F6gq9Gs. The second “The Dynasty of Asenevtsi” [19] game’s thematic topic is Bulgarian medieval history from the 12th and 13th centuries. This game comprises information from the following three time periods: the uprising of Peter and Assen, the ruling of King Kaloyan, and King Ivan Assen II. The maze includes four halls, the first one for introducing the game, and each of the following three is dedicated to one of the periods mentioned above. The player must go through the first three halls to reach the last one, solve the tasks, and answer the questions for unlocking the doors. The game ends after all the objects hidden in the halls are found. This game targets the audience of fifthand sixth-grade students, corresponding from 12 to 14 years old adolescents. A sample video is available at https://youtu.be/mI9NwiZOrB0. The next game is “The Legacy of Valchan Voyvoda” [24], which is dedicated to the life and work of the legendary Bulgarian Valchan voivode. The player passes through the maze halls by answering questions for unlocking the doors, solving various didactic tasks, and searching for hidden objects. This educational maze game contains seven halls with the following topics: 1. Introductory; 2. Age; 3. Who is Valchan Voyvoda; 4. Lifework; 5. The Legacy; 6. Valchan Voyvoda today; 7. Treasure. The game ends after the player visits every maze hall and finds all the hidden objects. A sample video is available at: https://youtu.be/QI5bplmxxZo. The fourth presented game, “Let Us Save Venice” [12], intends to educate viewers on the effects of climate change, especially as it relates to sea-level rise and cultural heritage destruction in areas like Venice [25]. To finish the game, the players must solve all required puzzles, pass through the maze, and detect the translucent objects. They can go to the next hall by answering a topical question correctly and earning the highest possible score. On the hall’s walls are boards with information concerning the didactic mini-games and questions. The maze contains five interconnected halls: Introduction, Context, The Problem, The Solution, and The Future. The game is targeted at students from a variety of specialities in different universities. The thematic area is significant and influences all national and age groups, and “Let Us Save Venice” is therefore suited for life-long learning and information activities. A recorded video with a sample game-play is available at https://youtu.be/YiBI2K9hC88.

4 Discussion All four educational video games outlined in the previous section are of maze type enriched by various 2D and 3D mini-games. The maze games have been constructed using the APOGEE platform to automate the creation of smart games for learning.

Ready to Play - A Comparison of Four Educational Maze Games

91

Table 1 presents a comparison of their main characteristics. All the games are generated by using the APOGEE software platform from XML formal descriptions that differ in language, number of maze halls, number of learning boards, and included mini-games (such as different types of mini-games and number of game instances embedded into maze rooms), audio resources and images incorporated in maze halls, and inclusion of NPC’s and automatic TTS (Text-To-Speech) translation. The four games also differ in the time needed for generating the maze in Unity 3D Editor (based on the XML formal description and archives with images and audio) and the time necessary for the game to be built as a desktop application or a browser WebGL game. Table 1. Characteristics of the developed video games of type enriched maze. Game

Feature

Carpet Industry in Bulgaria

The Dynasty of Asenetsi

The Legacy of Valchan Voyvoda

Let Us Save Venice

Language

Bulgarian

Bulgarian

Bulgarian

English

Number of halls

6

5

8

5

Number of learning boards M ini-games (types/instances)

48 3/9

34 6 / 13

54 5 / 12

38 5/8

Audio resources

6

5

8

5

Images

98

118

147

82

XM L (lines / kB)

801 / 32

1026 / 107

1203 / 154

771 / 102

NPCs included

No

Yes

No

No

TTS

No

Yes

No

No

5 sec

5 sec

M aze generation time

2 sec

Time for game build Distribution

3.5 min Desktop

6.2 / 14.6 min 7.9 / 18.3 min Desktop/Online Desktop/Online

4 sec 12.3 min Online

With increasing the number of game resources (text, images, and audio files) and the size of the XML document, both these times increase. However, the time needed for maze generation remains relatively shorter than the time for a game build (desktop or online). Nevertheless, the time for maze generation and the one for the game build are much shorter than the time for collecting and structuring learning content, which may be extended up to several weeks or months. Provided that game designers have the didactic content ready (e.g., structured in a lesson or a textbook), they have to create an XML formal description, either by using the APOGEE graphic editor for maze creation and further validation and generation of the XML file, or manually through an XML template for describing enriched maze games. The XML document should be validated using an XML Schema definition of enriched maze games. Next, this XML document is given to the APOGEE Maze Builder, which generates the maze game in the Unity 3D editor. The generated maze game can be immediately built employing the Unity Builder interface; however, some minor modifications need to be done, such as relocations of game objects, changes in illumination, etc. The game can be distributed through the players by sending

92

E. Paunova-Hubenova et al.

them the executable files (for desktop games) or the Internet (for any game). Next, the games can be played without restrictions or in a controlled mode. The last is applied for game validation and includes pre-game quizzes (e.g., for acquiring demographic characteristics of the players, their previous gaming experience and knowledge in the domain, learning style [26] and playing style [27], etc.), game sessions with tracking of player performance and outcomes, and post-game quizzes for assessing game experience [28] together with playability and learnability [29]. All four games have been validated successfully, having excellent results in-game experience, playability, and learnability [12, 14, 23, 24].

5 Conclusion and Future Work The paper presented four educational video games constructed using the APOGEE game platform. The first game (“Carpet industry in Bulgaria”) is focused on the manufacture and industrial development of the carpet industry in Bulgaria; the second one (“Let Us Save Venice”) is dedicated to climate resilience of built cultural heritage and the last two games are about Bulgarian history namely about the legacy of the legendary Valchan Voyvoda (who has lived in 18–19 century AD) and about the dynasty of Asenevtsi during the Second Bulgarian Kingdom (12–13 century AD). This comparison is limited only to some game characteristics (Table 1), but the authors plan to evaluate the four games through practical experiments with target users. The game platform APOGEE has been validated as a promising instrument and methodology for designing and developing learning games adapted for multiple educational scenarios. The maze provides an easy-to-understand structure of the multimedia learning content, including different challenges presented as mini-games and puzzles. The future steps cover wider dissemination among target audiences, attracting teachers and experts, and even learners and students to design and deploy various educational games. Further efforts will be made to improve the mini-games’ personalization and adaptation possibilities, learning analytics, and smart services for delivering more enjoyable, efficient, and learner-oriented educational experiences. Acknowledgements. The research is partially supported by the APOGEE project, funded by the Bulgarian National Science Fund, No. DN12/7/2017.

References s 1. Zhonggen, Y.: A Meta-analysis of use of serious games in education over a decade. Int. J. Comput. Games Technol. 2019(Article ID 4797032), 8 (2019) 2. Mouaheb, H., Fahli, A., Moussetad, M., Eljamali, S.: The serious game: what educational benefits? Procedia Soc. Behav. Sci. 46, 5502–5508 (2012) 3. Allisop, Y., Yildirim, E., Scepanti, M.: Teachers’ beliefs about game based learning: a comparative study of pedagogy, curriculum and practice in Italy, Turkey and the UK. In: Proceedings of ECGBL 2013, pp. 1–10 (2013) 4. Connolly, T.M., Hainey, T., Boyle, E., Baxter, G., Moreno-Ger, P. (eds.): Psychology, Pedagogy, and Assessment in Serious Games. Information Science Reference (2014)

Ready to Play - A Comparison of Four Educational Maze Games

93

5. Zirawaga, V.S., Olusanya, A.I., Maduku, T.: Gaming in education: using games as a support tool to teach history. J. Educ. Pract. 8(15), 55–64 (2017) 6. Dicheva, D., Dichev, C., Agre, G., Angelova, G.: Gamification in education: a systematic mapping study. Educ. Technol. Soc. 18, 75–88 (2015) 7. Keremedchiev, D., Borissova, D., Tuparov, G.: An algorithm for assessment of students using gamification. In: Proceedings of International Convention (MIPRO), pp. 636–640 (2020) 8. Bontchev, B.: Modern trends in the automatic generation of content for video games. Serdica J. Comput. 10(2), 133–166 (2016) 9. Terzieva, V., Paunova-Hubenova, E., Bontchev, B., Vassileva, D.: Teachers need platforms for construction of educational video games. In: Proceeding of EDULEARN, pp. 8260–8270 (2018) 10. Terzieva, V., Paunova-Hubenova, E., Bontchev, B.: Identifying the user needs of educational video games in Bulgarian schools. In: Proceedings of ECGBL 2018, pp. 687–695 (2018) 11. Guillermo, V.-G., Naipe, D.: Clover quiz: a mobile trivia game based on DBpedia data. In: International Semantic Web Conference (2017) 12. Bontchev, B., Antonova, A., Terzieva, V., Dankov, Y.: “Let Us Save Venice”—an educational online maze game for climate resilience. Sustainability, MDPI 14, 7 (2022) 13. Dankov, Y., Bontchev, B.: Towards a taxonomy of instruments for facilitated design and evaluation of video games for education. In: Proceedings of CompSysTech 2020, pp 285–292 (2020) 14. Paunova-Hubenova, E.: Didactic mini video games – students’ and teachers’ point of view, In: CBU International Conference Proceedings, vol. 7, pp. 552–558 (2019) 15. Andreeva, A.: Conceptual Requirements for non-traditional exhibition spaces. Architectural, design and art aspects. Bulgarian Journal for Engineering Design, Issue 40, Mechanical Engineering Faculty, TU-Sofia, pp. 53–58 (2019). ISSN:1313-7530 16. Deykov, Y., Andreeva, A.: Current aspects of the virtual design of expo-environment - gallery, Museum, Church. In: International Conference Modern Technology in Cultural Heritage, vol. 5, TU-Sofia, pp. 17–22 (2017). ISSN:2367–6523 17. Andreeva, A.: Colorful and General-Artistic Aspects of Architecture and Design. Viewpoints. Aesthetic Achievements of the Exhibition Activities of TU-Sofia 2009_2019, vol. 1, Issue 1, Technical University-Sofia, pp. 78–96 (2019). ISSN: 2682-9797 18. APOGEE Project site. https://apogee.online/. Accessed on 5 May 2022 19. Dankov, Y., Bontchev, B., Terzieva, V.: Design and creation of educational video games using assistive software instruments. In: Ahram, T.Z., Karwowski, W., Kalra, J. (eds.) AHFE 2021. LNNS, vol. 271, pp. 341–349. Springer, Cham (2021). https://doi.org/10.1007/978-3-03080624-8_42 20. Bontchev, B., Antonova, A., Dankov, Y.: Educational video game design using personalized learning scenarios. In: Gervasi, O., et al. (eds.) ICCSA 2020. LNCS, vol. 12254, pp. 829–845. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-58817-5_59 21. Bontchev, B., Terzieva, V., Paunova-Hubenova, E.: Personalization of serious games for learning. Interact. Technol. Smart Educ. 18(1), Emerald (2021) 22. Warpefelt, H., Verhagen, H.: A model of non-player character believability. J. Gam. Virtual Worlds 9(1), 39–53 (2017) 23. Bontchev, B., Panayotova, R.: Towards automatic generation of serious maze games for education. Serdica J. Comput. 11(3), 249–278 (2017) 24. Bontchev, B., Terzieva, V., Dankov, Y.: Educational video maze games. Nauka J. SAI XXXI(1), 25–33 (in Bulgarian) (2021) 25. Taramelli, A., Valentini, E., Piedelobo, L., Righini, M.: Assessment of state transition dynamics of coastal wetlands in northern Venice lagoon. Italy. Sustain. 13, 4102 (2021) 26. Soflano, M., Connolly, T.M., Hainey, T.: An application of adaptive games-based learning based on learning style to teach SQL. Comput. Educ. 86, 192–211 (2015)

94

E. Paunova-Hubenova et al.

27. Bakkes, S.C., Spronck, P.H., van Lankveld, G.: Player behavioural modelling for video games. Entertain. Comput. 3(3), 71–79 (2012) 28. Calvillo-Gámez, E.H., Cairns, P., Cox, A.L.: Assessing the core elements of the gaming experience. In: Bernhaupt, R. (ed.) Game user experience evaluation. HIS, pp. 37–62. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-15985-0_3 29. Sánchez, J.L.G., Vela, F.L.G., Simarro, F.M., Padilla-Zea, N.: Playability: analysing user experience in video games. Behav. Inform. Tech. 31(10), 1033–1054 (2012)

Employing FFNN and Learning Styles to Improve Knowledge Acquisition in Educational Digital Games Christos Troussas(B)

, Akrivi Krouska , and Cleo Sgouropoulou

Department of Informatics and Computer Engineering, University of West Attica, 12243 Egaleo, Greece {ctrouss,akrouska,csgouro}@uniwa.gr

Abstract. This paper presents an educational digital game which combines FFNN and learning styles to provide optimal types of learning units to learners so they can advance their knowledge level and improve their and learning outcomes. The digital game has ten stages and its main scope it to teach the players basic computer science concepts. In each stage, there are agents who give the learning units to the player in the gamified environment. The learning activities are delivered using a Feed Forward Neural Network (FFNN) and Weighted Sum Model (WSM) for optimizing the types of delivered learning activities to students based on their learning style. The students’ learning style is based on the Honey and Mumford model. The final output of the FFNN is different types of learning activities, delivered to students. For the evaluation of the game, we used a questionnaire and the statistical hypothesis test. The results show high level of acceptance of the presented model in digital games environment. Keywords: Educational digital game · ANN · WSM · Learning activities · Intelligent tutoring system · Adaptive learning

1 Introduction The field of education is turning into a really overbearing and complex field throughout the years with instructors attempting to convey more effective approaches to rendering traditional or online classes an environment where knowledge can be better transmitted. Traditional learning strategies hinder many burdens for students in terms of understanding and absorbing new concepts. Instructors have investigated and utilized a wide range of strategies and/or technologies to promote education, for example, social networks [1– 5], smartphones and handheld devices [6–8], Mixed Reality [9–11], and video games [12–14]. Video games give another crisp, energizing and viable strategy for learning, however at times that may not be enough for learners to remain motivated during the educational process. They provide a pleasant and motivating environment that can serve as an optimal solution for students to learn in it. However, video games are played by different players. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 A. Krouska et al. (Eds.): NiDS 2022, LNNS 556, pp. 95–103, 2023. https://doi.org/10.1007/978-3-031-17601-2_10

96

C. Troussas et al.

The players may have different knowledge, different pace of learning and different preferences to learn and be evaluated. The incorporation of adaptivity [15, 16] in video games can be a solution to the above issue and a powerful idea to promote education. One mechanism of adaptivity is learning styles. Learning styles involve a scope of hypotheses that intend to represent differentiations in people’s learning. Researchers share the recommendation that people can be driven by their “style” of learning, however there are differences in how these proposed “styles” ought to be characterized, sorted and evaluated. A typical idea is that people differ in the way they learn. There are many examples of learning styles, e.g. Kolb’s model, Honey and Mumford’s model, Felder and Silverman learning style assessment, etc. [17]. Artificial Neural Networks (ANNs) are computing systems which intend to reproduce the manner in which the human mind dissects and processes data. The innovation of ANNs has been used by researchers in the field of learning technology to offer personalization to learners’ needs and preferences [18, 19]. FFNN is a type of ANN, in which connections between the nodes do not form a cycle. In FFNN, data moves in just a single route – forward – from the input nodes, through the hidden nodes and to the output nodes. Analyzing the related literature, ANNs have been used in learning technology systems [18–26] to provide a tailored learning experience to students. For example, in [18– 20], the authors have emphasized on discovering similarities concerning the knowledge patterns between the students’ profiles and the learning units. In [21, 22], the adaptation of the tutoring system to the learners’ needs and preferences using ANNs has been investigated. In [23], the researchers have used ANN to build a recommendation engine to held learners during the educational procedure. Other attempts have concentrated on the delivery of adaptive education using ANN through the formation of learning routes that are tailored to the students’ wishes and skills [24, 25]. Finally, ANNs have been also used for emotion analysis in tutoring systems [18, 26]. In view of the above, this paper presents a digital game which incorporates FFNN and learning styles to deliver optimal learning activities to players so that they can further advance their knowledge level. The digital game has ten different stages and its main scope it to teach the players basic computer science concepts, such as Algorithms and Tractability, Programming Languages and Compilers, etc. In every stage, there are agents who deliver learning activities to the player. Hence, the player is offered a plethora of learning activities designed to advance his/her basic concepts of computer science. These learning activities are delivered using ANN and weighted Sum Model for optimizing the learning activities to students based on their learning style. The students’ learning style is based on the Honey & Mumford model. The final output of the ANN is different types of learning activities in which learners can participate. For the evaluation of the system, we used a questionnaire and the statistical hypothesis test. The results show high level of acceptance of the presented model in digital games environment.

2 Learning Styles in the Video Game The learning style describes the manner with which an individual learns. The learning styles of students are set through a questionnaire, being conveyed to them during their

Employing FFNN and Learning Styles

97

registration in the system. In our methodology, we use the Honey and Mumford learning style model [27] to deliver learning units, adapted to the video game environment, with the aim to further develop the learning results. The motivation behind why we utilized the Honey and Mumford model is that includes learning approaches that people normally like. In the Honey and Mumford model, learning is amplified when learners comprehend how they can really advance their knowledge. Besides, this model gives an exceptionally refined self-discernment inventory being fundamental for students to figure out their predominant learning style. The attributes of the four learning styles proposed by this model are summed up as follows: • Activists (AC): Activists are learners who learn by acting. Hence, they learn better through learning activities, such as brainstorming, problem-solving, group discussion, puzzles, competitions, role-play etc. • Theorists (TH): Theorists need to delve into the theory behind the exercises in order to learn better. The learning activities they prefer include models, statistics, stories, quotes, background information, applying concepts theoretically etc. • Pragmatists (PR): These learners are practical inclining towards applying the new ideas to problems of the real life. The learning activities, facilitating their knowledge acquisition, include experimenters, trying out new ideas, case studies, discussion etc. • Reflectors (RE): These people learn by observing and reflecting on outcomes. They prefer to examine different perspectives, collect data and afterwards work towards a conclusion. Appropriate learning activities for reflectors are observing activities, feedback from others, paired discussions etc.

3 Adaptive Learning Units Using FFNN For the types of learning units, delivered to players for optimal knowledge acquisition, the system incorporated ANN blended with the Weighted Sum Model (WSM), being one of the most efficient Multiple Criteria Decisions Analysis tools [28]. The reason why this technique is embraced is due to the fact that it renders the system dynamic and robust. The proposed activities are fully incorporated into the digital game. The learning style of each student based on the Honey & Mumford model is the input of the ANN. Specifically, for each type of the learning style (AC, TH, PR, RE), different weights are assigned. Practically, this means that a student prefers to receive a specific learning activity based on his/her learning style but also s/he may prefer other activities (belonging predominantly to different learning style) in a different percentage. The other types of the learning activities are given the corresponding weights following the same rationale. Then, the activation function, which is calculated using WSM, gives the output of the FFNN by mapping the resulting values, as shown in Fig. 1. The weights and the activities of the ANN the loads in the ANN have been determined by 10 experts in the field of education, who are university professors and teachers in primary and secondary education (6 university professors and 4 teachers). The experts hold a PhD degree and have more than 10 years of experience and involvement in education. The output of the ANN is a different percentage of learning activities being assigned to each learner based on his/her learning style, as follows:

98

C. Troussas et al.

Z1 = (PAC W1AC + PPR W1PR ) Z2 = (PTH W2TH + PPR W2PR + PRE W2RE ) Z3 = (PAC W3AC + PRE W3RE ) Z4 = (PTH W4TH + PPR W4PR ) Z5 = (PAC W5AC + PPR W5PR + PRE W5RE ) It needs to be emphasized that the learning activities are well-built into the game environment.

Fig. 1. ANN architecture.

To better illustrate the functionality of the ANN, the case of a student, named Panagiotis, is described. Panagiotis is 70% Activist and 30% Pragmatist, based on the system’s log files. As such, PAC = 0.70, PTH = 0.0, PPR = 0.30 and PRE = 0.0. The corresponding weights receive the following values: W1AC = 0.8, W3AC = 0.1, W5AC = 0.1, W1PR = 0.6, W2PR = 0.1, W4PR = 0.1 and W5PR = 0.2. These weights are the input to the activation function, which has the following outputs: Z1 = 0.7 * 0.8 + 0.3 * 0.6 = 0.74 Z2 = 0 + 0.3 * 0.1 + 0 = 0.03 Z3 = 0.7 * 0.1 + 0 = 0.07 Z4 = 0 + 0.3 * 0.1 = 0.03 Z5 = 0.7 * 0.1 + 0.3 * 0.2 + 0 = 0.13 Based on the output of the FFNN, the game delivers mainly critical thinking and paired conversation activities to players towards the improvement of their knowledge.

4 Evaluation Results For the evaluation of the presented educational digital game, we used the Lynch & Ghergulescu framework [29], which is oriented to the evaluation of adaptive and intelligent tutoring environments. This framework includes the following dimensions: 1) learning

Employing FFNN and Learning Styles

99

and training, 2) system, 3) user experience and 4) affective dimension. The first dimension (“learning and training”) assesses the effectiveness and productivity of the learning process and the knowledge acquisition of students. The second dimensions (“system”) assesses the efficiency of the application/software, i.e. the algorithmic techniques that are used to assist learners advance their knowledge. The third dimension (“user experience”) assesses the attitude of students towards using the system. The fourth dimension (“affective”) assesses the engagement of students during their interaction with the software. In the evaluation, the population consisted of 40 students of the Department of Informatics and Computer Engineering of a public university. All the participants are postgraduate students of the Master-level conversion course in Information Technology and Applications. The duration of the evaluation was 13 weeks. After this period, the students were asked to complete questionnaires. These questionnaires were based on the Lynch-Ghergulescu framework and included twelve questions, following a 1–10 ranking model (1 is lower and 10 is higher), as shown in Table 1. Table 1. Questionnaire. No

Question

Dimension

1

Rate your learning outcome improvement

Learning and Training

2

How efficient is the use of time?

3

Rate the type of delivered units for learning as well as their relevance to your learning style

4

Rate the adequacy of stages of the game

5

Rate your satisfaction

6

Rate your overall experience

7

Rate the easiness of use of the game

8

Rate the familiarity of the game

9

Rate the quality of the game

10

Rate the usefulness of the game

11

Rate your engagement with the game

12

How motivating is your learning experience?

System

User experience

Affective dimension

The answers of the students in the questions of Table 1 were aggregated based on the dimensions of the framework and are presented in Fig. 2. Analyzing the evaluation results, it can be inferred that the presented digital game can offer an efficient environment for optimal knowledge acquisition, since it provides learning activities to students in a successful way. To achieve this, it incorporates an intelligent mechanism with ANN and WSM, blended with the Honey and Mumford learning style instrument. In terms of the dimensions of the evaluation framework, the presented learning strategy, being incorporated in the digital game, can:

100

C. Troussas et al.

90%

80%

77.50%

80%

75%

72.50%

70% 60% 50% 40% 30% 20% 10%

22,5%

17.50%

15% 5%

5%

2.50%

25.00%

2.50%

0% Learning and Training

System

High

User experience

Average

Affecve

Low

Fig. 2. Questionnaire results.

i. help student better improve their learning outcomes (dimension 1):. ii. be effective for students, since it incorporates adequate algorithmic techniques to be implemented (dimension 2). iii. offer a pleasant experience to learners (dimension 3). iv. be engaging and motivating for the learners (dimension 4). Towards acquiring more qualitative results, we also used t-test. As such, the presented game was compared to its conventional version. The conventional version was the same game, with the same stages as well as the same incorporated types of learning units; however, the learning units were not delivered with the same sophisticated technique (using ANN and WSM), but in a random way. The presented game was used by 40 students (Group A), as mentioned earlier, while its conventional version was used by other 40 students (Group B). The two groups share the same characteristics. Since our research is focused mainly on the students’ learning outcomes and the system’s incorporated techniques, we applied t-test on Questions 1 and 3. Table 2 shows the t-test results. Considering the results of Table 2, there is a statistically significant difference between the means of the two trials concerning Questions 1 and 3. That means that the presented game surpasses its conventional version in terms of students’ learning outcome improvement and the incorporated sophisticated techniques for selecting the types of delivered learning units. The results were expected since our digital game, which is described in this paper, incorporates an intelligent mechanism (using ANN, WSM and a learning style instrument) for delivering the appropriate type of units for learning, being appropriately incorporated in the game environment.

Employing FFNN and Learning Styles

101

Table 2. T-test results. Metric

Question 1 Group A

Question 3 Group B

Group A

Group B

Mean

8,475

5,075

7,975

6,1

Variance

2,922436

3,404487

3,666026

2,861538

Observations

40

40

40

40

Hypothesized mean difference

0

0

df

78

77

t Stat

8,548947

4,641472

P(T