Intelligent Systems and Learning Data Analytics in Online Education

Description

Intelligent Systems and Learning Data Analytics in Online Education provides novel artificial intelligence (AI) and analytics-based methods to improve online teaching and learning. This book addresses key problems such as attrition and lack of engagement in MOOCs and online learning in general.

This book explores the state of the art of artificial intelligence, software tools and innovative learning strategies to provide better understanding and solutions to the various challenges of current e-learning in general and MOOC education. In particular, Intelligent Systems and Learning Data Analytics in Online Education shares stimulating theoretical and practical research from leading international experts. This publication provides useful references for educational institutions, industry, academic researchers, professionals, developers, and practitioners to evaluate and apply.

Presents the application of innovative AI techniques to collaborative learning activities
Offers strategies to provide automatic and effective tutoring to students’ activities
Offers methods to collect, analyze and correctly visualize learning data in educational environments

Front Cover 2
Intelligent Systems and Learning Data Analytics in Online Education 5
Copyright Page 6
Dedication 7
Contents 9
List of contributors 19
Foreword 23
Preface 27
- Final words 35
Acknowledgments 37
Conversational agents in MOOCs: reflections on first outcomes of the colMOOC project 39
- 1 Introduction 39
- 2 Background 40
  - 2.1 MOOCs and the social dimension of learning 40
  - 2.2 Conversational agents 41

Front Cover 2
Intelligent Systems and Learning Data Analytics in Online Education 5
Copyright Page 6
Dedication 7
Contents 9
List of contributors 19
Foreword 23
Preface 27
- Final words 35
Acknowledgments 37
Conversational agents in MOOCs: reflections on first outcomes of the colMOOC project 39
- 1 Introduction 39
- 2 Background 40
  - 2.1 MOOCs and the social dimension of learning 40
  - 2.2 Conversational agents 41
- 3 The colMOOC project 42
  - 3.1 Pedagogical rationale 42
  - 3.2 Agent design 42
  - 3.3 The colMOOC system architecture 44
  - 3.4 The colMOOC editor 44
  - 3.5 The colMOOC player 45
  - 3.6 Project research objectives and pilot MOOCs 47
- 4 “Computational Thinking” MOOC 48
  - 4.1 MOOC design 48
  - 4.2 Data collection and results 49
    - 4.2.2 The chat activity 49
    - 4.2.1 Participants 49
    - 4.2.3 Course and chat acceptance 50
  - 4.3 Reflections 51
- 5 “Programming for non-programmers” MOOC 52
  - 5.1 MOOC design 52
  - 5.2 Data collection and results 53
    - 5.2.2 Posttask questionnaires 53
    - 5.2.1 Course-entry questionnaire 53
    - 5.2.4 Course analytics 55
    - 5.2.3 Course-exit questionnaire 55
  - 5.3 Reflections 56
- 6 “Educational technology for the classroom” MOOC 57
  - 6.1 MOOC design 57
    - 6.1.3 Content 57
    - 6.1.1 Overview 57
    - 6.1.2 Participants 57
  - 6.2 Data collection and results 58
  - 6.3 Reflections 61
- 7 “Educational technologies in support for collaboration and evaluation in virtual learning environments” MOOC 62
  - 7.1 MOOC design 62
    - 7.1.1 Conversational agent activities design 62
  - 7.2 Data collection and results 63
    - 7.2.2 Apparatus and stimuli 64
    - 7.2.1 Participants 64
    - 7.2.3 MOOC evaluation results 65
    - 7.2.4 CA evaluation results 66
  - 7.3 Reflections 67
    - 7.3.2 CA activities 68
    - 7.3.1 MOOC course 68
    - 7.3.3 Final reflections 69
  - 8 Discussion and conclusions 69
  - Acknowledgments and disclaimer 74
  - References 74
  - Further reading 76
- PART I: Intelligent Agent Systems and Learning Data Analytics 77
  - 1 Artificial Intelligence (AI)-enabled remote learning and teaching using Pedagogical Conversational Agents and Learning An... 79
    - 1.1 Introduction 79
    - 1.2 Literature review 81
      - 1.2.1 Challenges and disappointments of online learning or remote teaching 81
      - 1.2.2 Artificial Intelligence, Pedagogical Conversational Agents, and Learning Analytics in support for remote teaching-lea... 82
    - 1.3 Our experience with Pedagogical Conversational Agents 84
      - 1.3.1 Building VIRTA in Unity3D 85
      - 1.3.2 VIRTA architecture and deployment environment 87
      - 1.3.3 Using the CA-VIRTA 87
    - 1.4 Results and analysis 90
      - 1.4.1 Access to VIRTA by teaching week 92
      - 1.4.2 Hours of access to VIRTA 92
      - 1.4.3 Access to VIRTA by week and hour 93
      - 1.4.4 Access to VIRTA by day of week 94
      - 1.4.5 Student interactions with VIRTA 96
    - 1.5 Discussion 97
      - 1.5.1 Utilizing Pedagogical Conversational Agents with Learning Management Systems 97
      - 1.5.2 Integrating Learning Analytics and Pedagogical Conversational Agents 99
    - 1.6 Future trends and conclusion 100
    - References 102
  - 2 Integrating a conversational pedagogical agent into the instructional activities of a Massive Open Online Course 107
    - 2.1 Introduction 107
    - 2.2 Literature review 108
    - 2.3 Research methodology 110
      - 2.3.1 Research goals and hypotheses 113
        
        2.3.1.1 Participants 114
        
        2.3.1.2 Instruments 114
      - 2.4 Results and discussion 115
        
        2.4.1 Discussion 116
      - 2.5 Conclusions and future work 118
      - References 119
    - 3 Improving MOOCs experience using Learning Analytics and Intelligent Conversational Agent 123
      - 3.1 Introduction 123
      - 3.2 Online learning and MOOCs 125
      - 3.3 Methodology 129
      - 3.4 LAICA integration in MOOCs: framework 130
        
        3.4.1 Proposed framework 130
        
        3.4.2 LAICA: system flow 132
        
        3.4.3 LAICA: analytics and dissemination 133
      - 3.5 LAICA integration in MOOCs: example implementation 134
        
        3.5.1 Knowledge base creation 134
        
        3.5.2 Database integration 137
        
        3.5.3 Test plan 138
        
        3.5.4 Learning analytics integration 138
      - 3.6 LAICA integration in MOOC: impact analysis 140
      - 3.7 LAICA integration in MOOC: findings and discussion 142
      - 3.8 Conclusion 144
      - References 145
      - Further reading 146
    - 4 Sequential engagement-based online learning analytics and prediction 147
      - 4.1 Introduction 147
      - 4.2 Related work 150
        
        4.2.1 Learning analytics 150
        
        4.2.2 Predicting student’s performance 151
        
        4.2.3 Students’ learning engagement 152
      - 4.3 Sequential engagement-based academic performance prediction network 153
        
        4.3.1 Notation and problem statement 153
        
        4.3.2 SPEN structure overview 155
        
        4.3.3 Engagement detector 156
        
        4.3.4 Sequential predictor 156
        
        4.3.5 Loss function 157
      - 4.4 Experiments and evaluation 157
        
        4.4.1 Experiments environment 157
        
        4.4.2 Experiment settings 157
        
        4.4.3 Compared methods and metrics 159
        
        4.4.4 Evaluation 160
        
        4.4.5 Experiment results and discussion 160
        
        4.4.5.1 Experiment A: impacts of historical performance 161
        
        4.4.5.2 Experiment B: impacts of engagement detector 162
        
        4.4.5.3 Experiment C: impacts of demographic information 162
        
        4.5 Conclusions and future work 163
        
        Acknowledgment 164
        
        References 164
      - 5 An intelligent system to support planning interactive learning segments in online education 167
        
        5.1 Introduction 167
        
        5.2 Theoretical backgrounds of intelligent systems in active learning systems 168
        
        5.3 Methodology of implementation and target group 169
        
        5.3.1 Learning objects in Moodle system 169
        
        5.3.2 Relationship learning object: learning topic 171
        
        5.3.3 Structure of observer as activity module 172
        
        5.3.5 Generalized sequential patterns of learning object 173
        
        5.3.4 Data analytics techniques 173
        
        5.3.6 Felder and Silverman learning style model 178
        
        5.4 Results and discussion 178
        
        5.5 Conclusion 183
        
        References 183
      - PART II: Artificial Intelligence Systems in Online Education 185
        
        6 A literature review on artificial intelligence and ethics in online learning 187
        
        6.1 Introduction and motivations 187
        
        6.2 Artificial intelligence in online learning 189
        
        6.2.1 Chatbots and educational bots 190
        
        6.3 Ethics in online learning 191
        
        6.4 Ethics in artificial intelligence 192
        
        6.4.1 Data science and ethics 194
        
        6.4.2 Opening the black box with explainable artificial intelligence 196
        
        6.5 Limitations of ethics by design and how moral systems can overcome them 197
        
        6.6 Reflections and guidelines for an ethical use of artificial intelligence in online learning 199
        
        6.6.1 Data science and fairness 200
        
        6.6.2 Transparency and honesty 200
        
        6.6.3 Purpose of intelligent system 201
        
        6.6.4 Technological exclusion and digital divide 202
        
        6.7 Concluding remarks and future work 203
        
        Acknowledgments 204
        
        References 204
        
        7 Transfer learning techniques for cross-domain analysis of posts in massive educational forums 209
        
        7.1 Introduction 209
        
        7.2 Related works 211
        
        7.3 Text categorization model 214
        
        7.3.1 Word-level document representation 215
        
        7.3.2 Overall document representation and classification 215
        
        7.4 Transfer learning strategy 218
        
        7.5 Experiments and evaluation 220
        
        7.5.1 Text categorization performance 222
        
        7.5.2 Transfer learning performance 223
        
        7.6 Conclusions and further work 225
        
        References 226
        
        8 Assisted education: Using predictive model to avoid school dropout in e-learning systems 229
        
        8.1 Introduction 229
        
        8.2 Background 232
        
        8.2.1 Recommender systems 232
        
        8.2.2 Predictive learning models 233
        
        8.2.3 Learning styles 235
        
        8.3 Related work 236
        
        8.3.1 Comparative analysis 237
        
        8.4 PRIOR Ensemble Architecture 238
        
        8.4.1 Prediction process 239
        
        8.5 DPE-PRIOR: Dropout predictive Ensemble Model 240
        
        8.5.1 DPE-PRIOR in action 243
        
        8.5.1.1 Experiment results and discussion 246
        
        8.6 Final remarks 248
        
        Acknowledgment 252
        
        References 252
        
        9 Adaptive task selection in automated educational software: a comparative study 255
        
        9.1 Introduction 255
        
        9.2 The pedagogical implications of personalized learning systems 256
        
        9.3 Overview of the different adaptation schemes 257
        
        9.3.1 Reinforcement learning methods 258
        
        9.3.1.1 Q-learning 258
        
        9.3.1.2 Temporal difference learning 259
        
        9.3.1.3 Virtual learning 259
        
        9.3.2 Bayesian inference-based methods 260
        
        9.3.2.1 Bayesian inference algorithm 260
        
        9.3.2.2 Bayesian knowledge tracing 261
        
        9.3.2.3 Bayesian networks 261
        
        9.3.2.4 Multiarmed bandit-based methods 262
        
        9.3.2.5 Item response theory and personalized testing methods 263
        
        9.3.3 Summary of the comparison of adaptive methods 264
        
        9.4 Methodology and research approach 266
        
        9.4.1 The evaluation functions used in our study 266
        
        9.4.2 A description of the simulated environment 267
        
        9.5 Simulation results 269
        
        9.5.1 Top performing algorithms 270
        
        9.5.3 Comparison of algorithms using different evaluation functions 271
        
        9.5.2 Comparison of algorithms over time 271
        
        9.5.4 Algorithm comparison for dynamically changed level of students 273
        
        9.6 Discussion 274
        
        9.7 Conclusions 276
        
        References 276
        
        Appendix A 279
        
        10 Actor’s knowledge massive identification in the learning management system 281
        
        10.1 Introduction 281
        
        10.2 Diagnosis of higher education in Morocco and contribution of e-learning 283
        
        10.2.3 Learner’s traces 284
        
        10.2.1 Position of e-learning in structuring information and communication technology projects in Morocco 284
        
        10.2.2 The interaction traces 284
        
        10.2.4 Tutor’s traces 285
        
        10.2.5 Administrator’s traces 286
        
        10.2.6 Researcher’s traces 286
        
        10.2.7 Learning management system traces files 287
        
        10.2.7.1 Traces in Open edX 287
        
        10.2.7.2 Traces in Moodle 287
        
        10.2.8 Trace management by big data 287
        
        10.2.9 Collect and loading traces process with big data tools 289
        
        10.2.10 The data storage 291
        
        10.2.11 Data analysis 292
        
        10.2.11.1 Batch processing MapReduce 293
        
        10.2.11.2 Machine learning in big data 294
        
        10.3 Machine learning evaluation tools 294
        
        10.3.1 Seeking information 295
        
        10.3.2 Data visualization tools 296
        
        10.4 E-learning massive data 296
        
        10.4.1 The analysis process of massive traces in a massive online open coursediscussion forum 297
        
        10.4.2 Extraction knowledge by big data analysis of massive log files 298
        
        10.4.3 Extraction knowledge by statistical descriptive analysis of log files 299
        
        10.4.4 Semantic analysis of big data in the discussion forum 305
        
        10.5 Conclusion 309
        
        References 309
        
        PART III: Applications of Intelligent Systems for Online Education 313
        
        11 Assessing students’ social and emotional competencies through graph analysis of emotional-enriched sociograms 315
        
        11.1 Introduction 315
        
        11.2 State-of-the-art analysis 317
        
        11.2.1 Sociometric assessment approaches 317
        
        11.2.1.1 Sociometry theory 317
        
        11.2.1.2 Sociometric assessment in educational groups 319
        
        11.2.1.3 Sociometric indexes 321
        
        11.2.1.4 Limitations of sociometric procedures 323
        
        11.2.2 Emotional intelligence and competences as social graph attributes 324
        
        11.2.2.2 Emotional intelligence assessment tools 325
        
        11.2.2.1 Emotional intelligence models 325
        
        11.2.2.3 Limitations of emotional intelligence models and assessment tools 328
        
        11.2.3 Motivation and adopted models 329
        
        11.3 Socioemotional graph analysis 330
        
        11.3.1 Data collection, processing, and emotional-enriched sociograms composition 332
        
        11.3.2 Social network analysis 335
        
        11.3.2.1 Detect influential members in a social group 336
        
        11.3.2.2 Create subgroups based on social distance across members 337
        
        11.3.2.3 Monitor the group’s social cohesion 338
        
        11.3.2.4 Link prediction between group members 339
        
        11.3.3 Provision of interactive recommendations for Social and Emotional Learning activities 340
        
        11.3.4 Indicative usage scenario in educational environment 342
        
        11.4 Application domains 343
        
        11.4.2 Work environments 343
        
        11.4.1 Educational environments 343
        
        11.4.3 Clinical environments 344
        
        11.5 Conclusions and open research areas 344
        
        References 345
        
        12 An intelligent distance learning framework: assessment-driven approach 349
        
        12.1 Introduction 349
        
        12.2 Sketch of intelligent distance teaching and learning framework 353
        
        12.3 Three-layered distance teaching and learning platform 354
        
        12.3.1 Module structure: Three Layers—Begin with Basic, Dive into Advanced, and Apply to Applications 356
        
        12.3.3 Answering questions posted by students on layers: Answer Collection and Ranking 357
        
        12.3.2 Layer structure: Four Panes—Slides, Videos, Summary, and Quizlet Panes 357
        
        12.4 Learning assessment and feedback 360
        
        12.4.1 Assessment question generation 361
        
        12.4.2 Smartphone assessment 363
        
        12.4.3 Question-driven reorganization of online course materials 365
        
        12.5 Analytics: mining from learning graph and assessment graph 367
        
        12.5.1 Assessment data analytics for distance teaching 369
        
        12.6 Simulation result 372
        
        12.7 Concluding remarks 373
        
        References 374
        
        13 Personalizing alternatives for diverse learner groups: readability tools 377
        
        13.1 Introduction 377
        
        13.2 Personalization in online education: trends, systems, and approaches 378
        
        13.2.1 Personalization and accessibility 380
        
        13.2.2 Readability as a starting point for personalization 381
        
        13.3 Implementing readability tools: general steps and suggestions 383
        
        13.3.1 Precourse preparation: content and display options 383
        
        13.3.2 Precourse preparation: content and tool selection 384
        
        13.3.3 Facilitating tool use—encouraging task crafting 385
        
        13.3.4 Evaluation of personalization and tool use 386
        
        13.3.5 Evaluation of performance 387
        
        13.3.6 Additional implementation considerations 387

Author's affiliation

Santi Caballé: Associate Professor of Computer Science, Open

Description

TOC

Reviews (0)

Author's affiliation