Handbook of educational data mining

Handbook of Educational Data Mining......Page 2
Chapman & Hall/CRCData Mining and Knowledge Discovery Series......Page 3
Handbook of Educational Data Mining......Page 4
Dedication\r......Page 6
Contents......Page 8
Preface......Page 12
Main Avenues of Research in Educational Data Mining......Page 13
\rEditors......Page 16
Contributors......Page 18
1.1 Background......Page 24
1.2 Educational Applications......Page 26
1.3 Objectives, Content, and How to Read This Book......Page 27
References......Page 28
Part I Basic Techniques, Surveysand Tutorials......Page 30
2.1 Introduction......Page 32
2.2.1 Visual Representations......Page 33
2.2.3 Abstract Data......Page 34
2.2.4 Cognitive Amplification......Page 35
2.3 Design Principles......Page 36
2.3.2 Graphical Excellence......Page 37
2.4.1.1 UM/ QV......Page 39
2.4.1.2 ViSMod......Page 40
2.4.1.3 E- KERMIT......Page 41
2.4.2.1 Simuligne......Page 42
2.4.3 Visualizations of Student- Tracking Data......Page 43
2.5 Conclusions......Page 47
References......Page 48
Contents......Page 50
3.1 Introduction......Page 51
3.2 Studies of Statistical Analysis of Web Log Files......Page 52
3.3.1 Log File Data......Page 53
3.3.2 Log File Data Abstractions......Page 54
3.4.1.3 Removal of Outliers......Page 55
3.4.4 User Identification......Page 56
3.5 Statistical Analysis of Log File Data......Page 57
3.5.1.1.1 Measure of Central Tendency......Page 58
3.5.1.2 Relationships between Variables......Page 59
3.5.2.1 Testing for Differences between Distributions Using Parametric Tests......Page 60
3.5.2.2 Testing for Differences between Distributions Using Nonparametric Tests......Page 61
3.6 Conclusions......Page 62
References......Page 63
4.1 Introduction......Page 66
4.2 The Pittsburgh Science of Learning Center DataShop......Page 67
4.3 Logging and Storage Methods......Page 68
4.5 Analysis and Visualization Tools......Page 72
4.6 Uses of the PSLC DataShop......Page 74
4.7 Data Annotation: A Key Upcoming Feature......Page 75
References......Page 76
5.1 Introduction......Page 80
5.2.1 Predicting Academic Success......Page 81
5.2.4 Metacognitive Skills, Habits, and Motivation......Page 82
5.3.1 Discriminative or Probabilistic Classifier?......Page 83
5.3.2 Classification Accuracy......Page 84
5.3.3 Overfitting......Page 85
5.3.4 Linear and Nonlinear Class Boundaries......Page 86
5.3.5 Data Preprocessing......Page 87
5.4.1 Decision Trees......Page 88
5.4.2 Bayesian Classifiers......Page 89
5.4.3 Neural Networks......Page 90
5.4.4 K-Nearest Neighbor Classifiers......Page 91
5.4.6 Linear Regression......Page 92
5.4.7 Comparison......Page 93
References......Page 94
6.1 Introduction......Page 98
6.2 The Clustering Problem in Data Mining......Page 99
6.2.1 k-Means Clustering......Page 101
6.2.2 Fuzzy c-Means Clustering......Page 102
6.2.3 Kohonen Self- Organizing Maps......Page 104
6.2.4 Generative Topographic Mapping......Page 105
6.3.1 Cluster Analysis of e- Learning Material......Page 106
6.3.2 Clustering of Students according to Their e- Learning Behavior......Page 108
6.4 Conclusions......Page 111
References......Page 112
7.1 Introduction......Page 116
7.2 Background......Page 117
7.3 Drawbacks of Applying Association Rule in e- Learning......Page 119
7.3.2 Discovering Too Many Rules......Page 120
7.3.3 Discovery of Poorly Understandable Rules......Page 121
7.3.4 Statistical Significance of Discovered Rules......Page 122
7.4 An Introduction to Association Rule Mining with Weka in a Moodle LMS......Page 123
References......Page 127
8.1 Introduction......Page 130
8.2.1 Background......Page 131
8.2.2 Tracing the Contextualized Learning Process......Page 132
8.3 Learning Log Analysis Using Data Mining Approach......Page 135
8.3.1 Preprocessing: From Events to Learning Actions......Page 136
8.3.2 Pattern Discovery......Page 137
8.3.3 Pattern Analysis: From Exploratory to Confirmatory Approach......Page 139
8.4 Educational Implications......Page 141
Acknowledgments......Page 142
References......Page 143
9.1 Introduction......Page 146
9.2 Process Mining and ProM Framework......Page 148
9.3 Process Mining Educational Data Set......Page 149
9.3.1 Data Preparation......Page 150
9.3.2 Visual Mining with Dotted Chart Analysis......Page 151
9.3.3 Conformance Analysis......Page 154
9.3.3.1 Conformance Checking......Page 155
9.3.3.2 LTL Analysis......Page 158
9.3.3.3 Process Discovery with Fuzzy Miner......Page 159
9.4 Discussion and Further Work......Page 161
References......Page 164
10.1 Introduction......Page 166
10.2.1 Conditional Independence and Marginal Dependence......Page 167
10.2.2 Nuisance Dependence......Page 171
10.3 Hierarchy......Page 173
10.3.2 Hierarchy of Institutional/ Social Structure......Page 174
10.4 Conclusions......Page 175
References......Page 176
Part II Case Studies......Page 180
11.1 Introduction......Page 182
11.2 Relation to Prior Work......Page 183
11.3.1 q-Matrix Algorithm......Page 185
11.3.2 Computing......Page 186
11.4.1 Binary Relations Tutorial, Section 1 (BRT-1)......Page 188
11.4.3 Binary Relations Tutorial, Section 3 (BRT-3)......Page 189
11.4.4 How Many Concepts and How Much Data?......Page 191
11.4.5 Summary of Expert-Extracted Comparison......Page 192
11.6 Conclusions......Page 193
References......Page 194
12.1 Introduction......Page 196
12.2 Theoretical Underpinning and Related Work......Page 197
12.3 Data......Page 198
12.4.1 Mirroring Visualizations......Page 199
12.4.2 Sequential Pattern Mining......Page 202
12.4.3.1 Clustering Groups......Page 203
12.4.3.2 Clustering Students......Page 204
12.5 Conclusions......Page 206
References......Page 207
13.1 Introduction......Page 210
13.2.1 Definition and Notation of Multi- Instance Learning......Page 211
13.2.2 Literature Review of Multi- Instance Learning......Page 212
13.3.1 Components of the Moodle Virtual Learning Platform......Page 214
13.3.2 Representation of Information for Working with Machine Learning Algorithms......Page 215
13.4 Experimentation and Results......Page 216
13.4.1 Problem Domain Used in Experimentation......Page 217
13.4.3 Comparison with Multi- Instance Learning......Page 218
13.4.4 Comparison between Single- and Multi- Instance Learning......Page 221
References......Page 222
14.1 Introduction......Page 225
14.2 Background......Page 226
14.3 Data......Page 227
14.4 Model......Page 228
14.5 Results......Page 231
14.6 Conclusions and Future Work......Page 233
References......Page 234
15.1 Introduction......Page 236
15.2 Related Work......Page 238
15.3 The AIspace CSP Applet Learning Environment......Page 239
15.4 Off- Line Clustering......Page 240
15.4.1 Data Collection and Preprocessing......Page 241
15.4.3.1 Cluster Analysis for the CSP Applet ( k = 2)......Page 242
15.4.3.2 Cluster Analysis for the CSP Applet ( k = 3)......Page 244
15.5 Online Recognition......Page 246
15.5.1 Model Evaluation (k=2)\r......Page 247
15.5.2 Model Evaluation for the CSP Applet (k=3)\r......Page 248
15.6 Conclusions and Future Work......Page 249
References......Page 250
16.1 Introduction......Page 254
16.2 Background......Page 255
16.3 Methodological Considerations......Page 256
16.4.1.2 Mining Dialogue Features from AutoTutor\'s Log Files......Page 257
16.4.2.1 Context......Page 258
16.4.2.2 Machine Learned Models from Student– System Interactions......Page 259
16.5 Discussion......Page 261
16.6 Conclusions......Page 263
References......Page 264
17.1 Introduction......Page 268
17.1.1.1 Association Rules and Associated Concepts......Page 269
17.1.1.2 Data from Logic- ITA......Page 270
17.1.1.3 Association Rules Obtained with Logic- ITA......Page 271
17.2.1 Some Measures of Interestingness......Page 272
17.2.2 How These Measures Perform on Our Datasets......Page 274
17.2.3 Contrast Rules......Page 276
17.3 Conclusions......Page 277
References......Page 278
18.1 Introduction......Page 280
18.2.1 Non- Multidimensional Paper Recommendation......Page 281
18.2.2 Contextual Recommendation with Multidimensional Nearest- Neighbor Approach......Page 282
18.3 Contextual Paper Recommendation with Multidimensional Nearest- Neighbor Approach......Page 284
18.4 Empirical Studies and Results......Page 287
18.4.2 Evaluation Results......Page 288
18.4.3 Discussions......Page 290
18.4.4 Implication of the Pedagogical Paper Recommender......Page 292
18.5 Concluding Remarks......Page 293
References......Page 294
19.1 Introduction......Page 296
19.2 Related Works......Page 297
19.3.1 Capturing Learning Experiences......Page 299
19.3.2 Learning Content- Based Profiles......Page 300
19.3.3 Detecting Active Interests......Page 303
19.3.4 Context- Aware Recommendation......Page 304
19.4 Case Study......Page 306
19.5 Conclusions......Page 307
References......Page 308
20.1 Introduction......Page 310
20.2 Motivation Measurement in Computer- Based Learning Configurations......Page 311
20.3.1 The Learning Environments......Page 312
20.3.5 Process......Page 313
20.3.6.1 Phase I— Constructing a Theory- Based Definition......Page 314
20.3.6.2 Phase II— Identifying Learning Variables......Page 315
20.3.6.4 Phase IV— Associating the Empirical Clusters with the Theory- Based Definition......Page 316
20.4 Discussion......Page 317
References......Page 318
21.1 Introduction......Page 322
21.2 Developing Scaffolding Capability: Mining Useful Information from Past Discussions......Page 323
21.2.2 Step 2: Technical Term Processing......Page 324
21.2.5 Step 5: Similarity Computation and Result Generation......Page 325
21.2.6 Step 6: Evaluation of System Responses......Page 326
21.3.1 Speech Act Classifiers......Page 327
21.3.2 Gender Classifier/ Distribution......Page 329
21.3.3 An Application of Gender Classifier/ Distribution......Page 330
21.4 Related Work......Page 331
References......Page 332
22.1 Introduction......Page 334
22.2 Context of the Study......Page 335
22.4.1 Abstraction Definitions......Page 336
22.4.3 Data Cleaning: Removal of Outliers......Page 337
22.5 Data Analysis Methods......Page 338
22.6.2.1 Sample Findings......Page 339
22.6.3.1 Sample Findings......Page 340
22.6.4.1 Sample Findings......Page 342
22.7 Discussion......Page 343
References......Page 345
23.1 Introduction......Page 346
23.2.1 Data Description......Page 347
23.2.2 Identifying Dependencies among Variables......Page 348
23.2.3 An Integrated Model of Behavior, Attitude, and Perceptions......Page 349
23.2.5 Case Study Summary......Page 352
23.3.1 Background and Related Work......Page 353
23.3.2 Data Description......Page 354
23.3.4 Students Express Their Emotions Physically......Page 355
23.4 Summary and Future Work......Page 358
Acknowledgments......Page 359
References......Page 360
24.1 Introduction......Page 362
24.2 Case Study: The UOC Digital Library......Page 363
24.3 Educational Data Analysis......Page 365
24.3.1 Data Acquisition......Page 366
24.3.2 UOC Users Session Data Set......Page 367
24.3.3 Descriptive Statistics......Page 368
24.3.4 Categorization of Learners\' Sessions......Page 369
References......Page 372
25.1 Introduction......Page 376
25.2.1 Problem Statement and Evaluation Criteria......Page 377
25.2.2 Anticipating Failure As Soon As Possible......Page 378
25.3.1 Problem Statement......Page 379
25.3.2 CAR- Based ASAP Classifiers......Page 380
25.4 Case Study......Page 382
References......Page 385
26.2 Motivation......Page 388
26.3 State of the Art......Page 390
26.4 The Key- Node Method......Page 391
26.5.2 Waikato Environment for Knowledge Analysis......Page 392
26.6.1 Data Description......Page 393
26.6.2 First Example......Page 394
26.6.3 Second Example......Page 395
Acknowledgment......Page 397
References......Page 398
27.1 Introduction......Page 400
27.3.1 Detection of Motivational Aspects in e- Learning......Page 401
27.3.3.1 Data Considerations......Page 402
27.3.3.2 Annotation of the Level of Engagement......Page 403
27.3.3.3 Analysis and Results......Page 404
27.3.3.4 Cross- System Results Comparison......Page 407
27.4 Challenges and Lessons Learned......Page 408
References......Page 409
Contents......Page 412
28.1 Introduction......Page 413
28.1.1 Relation to Prior Research......Page 415
28.1.2 Guidelines for Logging Tutorial Interactions......Page 416
28.1.2.3 Log Each School Year\'s Data to a Different Database......Page 417
28.1.2.6 Name Standard Fields Consistently Within and Across Databases......Page 418
28.1.2.10 Logging the Nonoccurrence of an Event Is Tricky......Page 419
28.2.1 Specify Events by When They Occurred......Page 420
28.2.2 Specify Events by a Database Query......Page 421
28.2.3 Specify Events by Their Similarity to Another Event......Page 423
28.3 Display Selected Events with the Context in Which They Occurred, in Adjustable Detail......Page 424
28.3.1.4 Duration and Hiatus......Page 425
28.3.2 Displaying the Event Tree......Page 426
28.3.2.2 Expanding the Event Tree......Page 427
28.4.1 Temporal Information......Page 428
28.4.4 Annotations......Page 429
28.5.1 Input Meta- Data to Describe Database Structure......Page 431
28.5.2 Which Events to Include......Page 432
28.5.3 Make Event Summaries Customizable by Making Them Queries......Page 433
28.5.4.2 Replicate Bugs......Page 434
28.6.1.2 Display Selected Events with the Context in Which They Occurred, in Dynamically Adjustable Detail......Page 435
28.6.2.3 Generality......Page 436
28.6.2.5 Utility......Page 437
References......Page 438
29.1 Introduction......Page 440
29.1.2 Background on the ASSISTment System......Page 441
29.2 Models: Creation of the Fine- Grained Skill Model......Page 442
29.2.1 How the Skill Mapping Was Used to Create a Bayesian Network......Page 443
29.3 Results......Page 444
29.4 Discussion and Conclusions......Page 446
Acknowledgments......Page 447
References......Page 448
30.1 Introduction......Page 450
30.3 Method......Page 451
30.4 Results......Page 455
30.4.1 Pair- Wise Analysis......Page 456
30.4.2 Entropy- Based Clustering......Page 458
30.5 Discussion......Page 460
References......Page 461
31.1 Introduction......Page 464
31.2.2 Datasets......Page 465
31.2.3 Employing Bayesian Networks for Student Modeling......Page 466
31.3.1 Predicting the Necessity of Help Requests......Page 467
31.3.2 Predicting the Benefit of Students\' Interactions......Page 469
31.4 Application and Future Work......Page 470
References......Page 472
32.1 Introduction......Page 474
32.2 The CanadarmTutor Tutoring System......Page 475
32.3 A Domain Knowledge Discovery Approach for the Acquisition of Domain Expertise......Page 476
32.3.1 Step 1: Recording Users\' Plans......Page 477
32.3.2 Step 2: Mining a Partial Task Model from Users\' Plans......Page 478
32.3.3 Step 3: Exploiting the Partial Task Model to Provide Relevant Tutoring Services......Page 480
32.3.3.3 Letting Learners Explore Different Ways of Solving Problems......Page 481
32.4 Evaluating the New Version of CanadarmTutor......Page 482
32.5.1 Other Automatic or Semiautomatic Approaches for Learning Domain Knowledge in ITS......Page 483
32.6 Conclusion......Page 485
References......Page 486
33.1 Introduction......Page 490
33.2 Background......Page 491
33.3 Creating an MDP- Tutor......Page 493
33.3.1 Constructing the MDP from Data......Page 494
33.3.2 The MDP Hint Generator......Page 496
33.4 Feasibility Studies......Page 497
33.5 Case Study: The Deep Thought Logic MDP- Tutor......Page 499
References......Page 501
34.1 Introduction......Page 504
34.3 Data Sources in Learning Objects......Page 505
34.3.2 External Assessments......Page 506
34.4 The Learning Object Management System AGORA......Page 507
34.5 Methodology......Page 508
34.5.2.2 Create Summarization Tables......Page 509
34.5.3.2 Classification Algorithms......Page 510
34.5.3.3 Association Algorithms......Page 512
References......Page 513
35.1 Introduction......Page 516
35.2 The Learning Style Model......Page 517
35.3 The Decision Model......Page 520
35.3.1 Building the Initial Model......Page 521
35.3.2 Adapting the Model......Page 522
35.5 Conclusions and Future Work......Page 523
References......Page 525