Resource-Adaptive Cognitive Processes (eBook)

Preface 6
Contents 8
Contributors 10
Resource-Adaptive Cognitive Processes 13
Jörg Siekmann and Matthew W. Crocker 13
1 Background 13
2 Resource-Adaptive Cognitive Processes 14
3 What Is a Resource-Adaptive Cognitive Process? 16
3.1 What Is a Resource-Limited Process? 16
3.2 How Can We Allocate the Given Resources to a Specific Task? 17
3.3 How Can We Allocate Resources That Are NotA Priori Defined? 17
3.4 Can We Postulate Indirectly Observable Resources That Help to Explain Experimental Data? 17
3.5 In a Multimodal Context, How Are Diverse Knowledge Resources Exploited? 18
4 Structure of This Volume 18
4.1 Part I: Resource-Bounded Cognitive Processes in Human Information Processing 18
4.2 Part II: Resource-Adaptive Processes in Human--Machine Interaction 20
4.3 Part III: Resource-Adaptive Rationality in Machines 21
References 21
Part I Resource-Bounded Cognitive Processes in Human Information Processing 23
Visuo-spatial Working Memory as a Limited Resource of Cognitive Processing 24
Hubert D. Zimmer, Stefan Münzer, and Katja Umla-Runge 24
1 The Concept of a Resource-Limited Working Memory 24
2 Components and Capacities of Visual Working Memory 27
3 Working Memory and Higher Cognitive Performances 32
4 Neural Structures Underlying Working Memory 34
5 Visual Working Memory in an Applied Context 37
References 40
From Resource-Adaptive Navigation Assistance to Augmented Cognition 46
Hubert D. Zimmer, Stefan Münzer, and Jörg Baus 46
1 Introduction 46
2 Resources 46
2.1 User's Resources 47
2.2 System's Resources 52
3 Goals 54
4 Assistance Systems of the Future: Augmented Cognition 57
References 60
Error-Induced Learning as a Resource-Adaptive Process in Young and Elderly Individuals 65
Nicola K. Ferdinand, Anja Weiten, Axel Mecklinger, and Jutta Kray 65
1 Introduction 65
2 Error Monitoring, ERN/Ne, and Dopamine 66
3 The Relevance of Learning Intention 68
4 Error-Induced Learning in the Elderly 70
5 Methods and Procedure 71
6 Results 73
6.1 Reaction Times 73
6.2 Error Rates 75
6.3 Speed-Accuracy Trade-Off 76
6.4 ERPs for Committed Errors 78
6.5 ERPs for Perceived Errors 79
7 Discussion 80
References 84
An ERP-Approach to Study Age Differences in Cognitive Control Processes 87
Jutta Kray and Ben Eppinger 87
1 Introduction 87
2 Cognitive Flexibility and Inhibition Limitations in Older Adults 88
2.1 Age Differences in Task Switching 88
2.2 Age Differences in Interference Control 89
2.3 Interactions Between Cognitive Control Processes and Their Temporal Dynamics 90
3 Methods Applied 91
3.1 Participants 91
3.2 Procedure 91
3.3 Data Recordings 92
4 Results 92
4.1 Behavioural Results 92
4.2 ERP-Results 94
4.2.1 Age Differences in ERP Correlates of Task-Preparation Processes 94
4.2.2 Age Differences in ERP Correlates of Interference Control 96
5 Summary and Conclusions 99
References 100
Simulating Statistical Power in Latent Growth Curve Modeling: A Strategy for Evaluating Age-Based Changes in Cognitive Resources 104
Timo von Oertzen, Paolo Ghisletta, and Ulman Lindenberger 104
1 Introduction 104
2 The Latent Growth Curve Model 106
3 Least Squares and Minus Two Log Likelihood Fitting Functions 108
3.1 Minimization of the Fitting Functions 109
3.2 Inadmissable Estimation Areas 112
4 General Simulation Procedure 114
4.1 Data Generation 114
4.2 Data Selection 115
4.3 Evaluation Criteria 116
4.4 Summarizing the Simulation Procedure 118
5 An Illustration 119
5.1 Population Parameters 119
5.2 Data Selection 120
5.3 Parameters of Focus 121
5.4 Definition of Power 121
5.5 Results 121
6 Discussion and Outlook 123
References 124
Conflicting Constraints in Resource-Adaptive Language Comprehension 127
Andrea Weber, Matthew W. Crocker, and Pia Knoeferle 127
1 Introduction 127
1.1 Incrementality 128
1.2 Multiple Constraints 128
1.3 Anticipation in Situated Comprehension 130
2 Varying Constraints 130
2.1 Discourse Information and Structural Preferences 131
2.2 Prosodic Information and Structural Preferences 134
2.3 Semantic Information and Lexical Preferences 137
2.4 Semantic Information and Visual Context 140
2.5 The Influence of the Scene: Depicted Events and Their Priority 142
3 Conclusions 146
References 147
The Evolution of a Connectionist Model of Situated Human Language Understanding 150
Marshall R. Mayberry and Matthew W. Crocker 150
1 Introduction 150
2 Experimental Findings 151
2.1 Anticipation in Unambiguous Utterances 152
2.2 Anticipation in Ambiguous Utterances 153
2.3 Coordinated Interplay Account 156
3 Connectionist Models 157
3.1 Multimodal Integration Using Event Layers 158
3.1.1 Input Data, Training, and Testing 159
3.1.2 Results 160
3.2 Multimodal Integration Using Attention 161
3.2.1 Input Data, Training, and Testing 162
3.2.2 Results 163
4 Conclusion 172
References 173
Part II Resource-Adaptive Processes in Human--Machine Interaction 175
Assessment of a User's Time Pressure and Cognitive Load on the Basis of Features of Speech 176
Anthony Jameson, Juergen Kiefer, Christian Müller, Barbara Großmann-Hutter, Frank Wittig, and Ralf Rummer 176
1 Introduction 176
1.1 Reasons for Variation in Cognitive Load and Time Pressure 176
1.2 Why Automatic Adaptation? 177
2 Possible Forms of Adaptation 178
2.1 Interruption of Communication 178
2.2 Timing and Form of Notifications 179
2.3 Dialog Strategy 179
2.4 Other Forms of Adaptation to Resource Limitations 180
3 Ways of Recognizing Resource Limitations 181
3.1 Recognizing Likely Causes of Resource Limitations 181
3.2 Physiological Indicators 181
3.2.1 Heart Rate Variability 182
3.2.2 Pupil Diameter 182
3.2.3 Other Indices 183
3.2.4 Comments 183
3.3 Evidence in the User's Behavior with the System 183
3.3.1 Evidence in the User's Motor Behavior 183
3.3.2 Evidence in the User's Speech 184
4 Experiments: Introduction 184
4.1 Earlier Research on Speech Indicators 184
4.1.1 Distinction from Other Topics 184
4.1.2 Effects of Cognitive Load 185
4.1.3 Effects of Time Pressure 185
5 Experimental Method 186
5.1 Purpose of Experiments 186
5.2 Method for Experiment 1 186
5.2.1 Materials 186
5.2.2 Design 186
5.2.3 Procedure 188
5.2.4 Subjects 188
5.2.5 Coding and Rating of Speech 188
5.3 Method for Experiment 2 189
6 Experimental Results 190
6.1 Statistical Analyses 190
6.2 Number of Syllables 190
6.3 Articulation Rate 191
6.4 Silent Pauses 192
6.5 Filled Pauses 193
6.6 Hesitations 194
6.7 Onset Latency 195
6.8 Disfluencies 195
6.9 Discussion 196
7 Learning of User Models 196
7.1 Bayesian Network Structure 197
7.2 Quantitative Parameters 199
7.3 Learning the Quantitative Parameters 200
8 Evaluation of the User Models 200
8.1 Procedure 200
8.2 Results 201
8.2.1 Recognizing Time Pressure 201
8.2.2 Recognizing Navigation 203
8.2.3 Dispensing with Individual Indicators 203
8.3 Discussion 204
9 Summary of Contributions and Remaining Work 206
References 206
The Shopping Experience of Tomorrow: Human-Centered and Resource-Adaptive 210
Wolfgang Wahlster, Michael Feld, Patrick Gebhard, Dominikus Heckmann, Ralf Jung, Michael Kruppa, Michael Schmitz, Lübomira Spassova, and Rainer Wasinger 210
1 Introduction 210
1.1 Overview Described Within a Motivating Scenario 211
2 Dialogue Shell of Talking Products 212
2.1 Modelling Personality in Voices 213
2.2 Expressing Personality in Dialogues 213
3 Mobile ShopAssist 215
4 Product Associated Displays 217
5 Personalized Ambient Soundscape Notification 218
5.1 Introduction to Ambient Audio Notification 219
5.2 Ambient Soundscapes and Audio Notification Cues 219
5.3 Applications and Shopping Scenario 221
6 Virtual Room Inhabitant 223
7 Live Acquisition of User Profile Data from Speech 227
8 Ubiquitous User Modeling with UbisWorld 231
9 Modeling Affect 234
9.1 Affect Taxonomy 234
9.2 Affect Computation 234
9.2.1 Mood Changes 236
9.2.2 Appraisal Based Affect Computation 238
10 Conclusions 238
References 239
Seamless Resource-Adaptive Navigation 243
Tim Schwartz, Christoph Stahl, Jörg Baus, and Wolfgang Wahlster 243
1 Introduction 243
2 REAL and BPN as the Basis of our Extensions 244
3 Overall System Architecture of the New Navigation Framework 245
4 Providing Map Material for Pedestrian Navigation 246
5 The Always Best Positioned Paradigm 251
5.1 Exocentric and Egocentric Localization 251
5.2 LORIOT 252
5.2.1 Estimation of the User Position 253
5.2.2 Orientation Estimation 256
5.2.3 Orientation Information Through Infrared Beacons 256
5.2.4 Orientation Information Through Active RFID Tags 257
5.2.5 Fusion of Orientation Information Through Bayesian Networks 257
5.2.6 Decomposition of a Direction Vector into Evidence Values 258
5.2.7 Composition of a Direction Vector out of Evidence Values 259
5.2.8 Example Calculation 259
6 Implementing a Seamless, Proactive User Interface 260
6.1 Hybrid Navigation Visualization 260
6.2 VISTO: Videos for Spatial Orientation 262
6.2.1 The Ubiquitous To-Do Organizer UBIDOO 263
6.2.2 The User Interface of VISTO 265
7 Summary 267
References 267
Linguistic Processing in a Mathematics Tutoring System: Cooperative Input Interpretation and Dialogue Modelling 270
Magdalena Wolska, Mark Buckley, Helmut Horacek, Ivana Kruijff-Korbayová, and Manfred Pinkal 270
1 Introduction 270
2 Research Setting 272
3 The Language of Informal Proofs 273
4 Baseline Processing 277
5 Aspects of Cooperative Interpretation 279
5.1 Parsing 280
5.2 Domain Modelling 281
5.3 Domain-Specific Anaphora 282
5.4 Flexible Formula Analysis and Disambiguation 283
6 Modelling Dialogue for Mathematics Tutoring 284
7 Related Work 287
8 Conclusions 288
References 289
Resource-Bounded Modelling and Analysis of Human-Level Interactive Proofs 293
Christoph Benzmüller, Marvin Schiller, and Jörg Siekmann 293
1 Introduction 293
2 The Need for Experiments and Corpora 295
3 Main Challenges and Resources for Proof Tutoring 298
3.1 B: Mathural Processing and Mathural Generation 298
3.2 C: Dialogue State and Proof Management 300
3.3 D: Proof Step Evaluation 300
3.4 E: Tutorial Context 301
3.5 F: Failure Analysis 302
3.6 G: Didactic Strategies, Feedback Generation and Hinting 302
3.7 H: Flexible Dialogue Modelling 302
4 Dynamic Proof Step Evaluation with MEGA 303
4.1 Proof Management, Correctness Analysis and Content Underspecification 303
4.2 Granularity Analysis 305
4.3 Learning Granularity Evaluation 305
4.4 Student Modelling 308
4.5 Further Work 309
5 Didactic Strategies and Dialogue Modelling 309
5.1 Didactic Strategies and Hinting 309
5.2 Dialog Modelling 310
6 Related Work and Conclusion 310
References 311
Part III Resource-Adaptive Rationality in Machines 314
Comparison of Machine Learning Techniques for Bayesian Networks for User-Adaptive Systems 315
Frank Wittig 315
1 Introduction: Bayesian Networks in User-Adaptive Systems 315
2 A Framework for Learning Bayesian Networks for User-Adaptive Systems 316
2.1 Machine Learning in User-Adaptive Systems 316
2.2 Learning Bayesian Networks for User-Adaptive Systems 317
2.2.1 Learning Offline (Batch) and Learning Online (Adaptation) 317
2.2.2 Exploiting Experimental and Usage Data for Learning 318
2.2.3 Learning Probabilities and Structure 318
2.2.4 Learning Interpretable Bayesian Network User Models 319
2.3 Learning Bayesian Network User Models in the READY Project 319
3 The Structural View: An Evaluation of Bayesian Networks User Model Learning 321
3.1 Combined Learning Approaches 321
3.2 Evaluation Procedure 323
3.3 Results 325
3.4 Discussion 328
4 Conclusion 329
References 334
Scope Underspecification with Tree Descriptions: Theory and Practice 337
Alexander Koller, Stefan Thater, and Manfred Pinkal 337
1 Introduction 337
2 Dominance-Based Scope Underspecification 338
2.1 Dominance Constraints 340
2.2 Dominance Graphs 341
2.3 Configurations and Solved Forms 342
3 Solving Dominance Constraints and Graphs 343
3.1 A Saturation Algorithm 345
3.2 Reduction to Set Constraints 346
3.3 A Graph-Based Solver 348
3.4 The Chart Solver 349
4 Practical Scope Underspecification 351
4.1 Minimal Recursion Semantics as Dominance Constraints 351
4.2 Experiments with the English Resource Grammar 353
4.2.1 Evaluating the Net Hypothesis 354
4.2.2 Grammar Verification 354
4.3 Redundancy Elimination 355
5 Annotating Scope 357
6 Conclusion 359
References 362
Dependency Grammar:Classification and Exploration 365
Ralph Debusmann and Marco Kuhlmann 365
1 Introduction 365
2 Dependency Structures 366
3 Dependency Structures and Lexicalized Grammars 368
3.1 Lexicalized Grammars Induce Dependency Structures 368
3.2 The Algebraic View on Dependency Structures 370
3.3 Regular Dependency Grammars 371
4 Extensible Dependency Grammar 373
4.1 Dependency Multigraphs 373
4.2 Grammars 374
5 Modeling Complex Word Order Phenomena 375
5.1 Scrambling 376
5.2 A Topological Model of Scrambling 376
6 A Relational Syntax--Semantics Interface 377
6.1 Dominance Constraints 378
6.2 The Interface 379
7 Modeling Regular Dependency Grammars 380
8 Grammar Development Environment 382
8.1 Parser 382
8.2 Large-Scale Parsing 383
9 Conclusion 385
References 385
OMEGA: Resource-Adaptive Processes in an Automated Reasoning System 389
Serge Autexier, Christoph Benzmüller, Dominik Dietrich, and Jörg Siekmann 389
1 Motivation and Historical Background 389
1.1 The OMEGA Initiative 392
2 Resource-Adaptive Proof Search 394
2.1 Human-Oriented High-Level Proofs 394
2.1.1 Inferences 395
2.1.2 Application Direction of an Inference 397
2.1.3 Representation of Proof 398
2.2 Searching for a Proof 400
2.2.1 Knowledge-Based Proof Search 400
2.2.2 Reactive Proof Search 403
3 Knowledge as a Resource 405
3.1 Managing Mathematical Knowledge 405
3.2 Formalising Mathematical Knowledge 406
3.3 From Assertions to Inferences 406
3.4 From Inferences to Planner Methods 408
3.5 From Inferences to Agents 409
4 Specialised Computing and Reasoning Resources 409
5 mega as an Adaptive Resource 411
5.1 Adaptation to Users with Different Skills 412
5.2 Adaptation to Different Software Systems 414
5.2.1 Checking the Correctness 415
5.2.2 Cognitive Proof States 416
6 Future Research 417
References 418