Mehr lesen
This book summarizes the main problems posed by the design of a man-machine dialogue system and offers ideas on how to continue along the path towards efficient, realistic and fluid communication between humans and machines.
A culmination of ten years of research, it is based on the author's development, investigation and experimentation covering a multitude of fields, including artificial intelligence, automated language processing, man-machine interfaces and notably multimodal or multimedia interfaces.
Contents
Part 1. Historical and Methodological Landmarks
1. An Assessment of the Evolution of Research and Systems.
2. Man-Machine Dialogue Fields.
3. The Development Stages of a Dialogue System.
4. Reusable System Architectures.
Part 2. Inputs Processing
5. Semantic Analyses and Representations.
6. Reference Resolution.
7. Dialogue Acts Recognition.
Part 3. System Behavior and Evaluation
8. A Few Dialogue Strategies.
9. Multimodal Output Management.
10. Multimodal Dialogue System Assessment.
About the Authors
Frédéric Landragin is a computer science engineer and has a PhD from the University of Lorraine, France. He is currently in charge of linguistics research for the French National Center for Scientific Research (CNRS). His studies focus on the analysis and modeling of language interpretation. Man-machine dialogue is one of the applications of this research.
Inhaltsverzeichnis
Preface xi
Introduction xv
PART 1 HISTORICAL AND METHODOLOGICAL LANDMARKS 1
Chapter 1 An Assessment of the Evolution of Research and Systems 3
1.1 A few essential historical landmarks 5
1.2 A list of possible abilities for a current system 16
1.3 The current challenges 23
1.4 Conclusion 27
Chapter 2 Man-Machine Dialogue Fields 29
2.1 Cognitive aspects 30
2.2 Linguistic aspects 40
2.3 Computer aspects 45
2.4 Conclusion 46
Chapter 3 The Development Stages of a Dialogue System 47
3.1 Comparing a few development progresses 48
3.2 Description of the main stages of development 52
3.3 Conclusion 62
Chapter 4 Reusable System Architectures 63
4.1 Run-time architectures 64
4.2 Design-time architectures 69
4.3 Conclusion 73
PART 2 INPUTS PROCESSING 75
Chapter 5 Semantic Analyses and Representations 77
5.1 Language in dialogue and in man-machine dialogue 78
5.2 Computational processes: from the signal to the meaning 85
5.3 Enriching meaning representation 91
5.4 Conclusion 94
Chapter 6 Reference Resolution 95
6.1 Object reference resolution 96
6.2 Action reference resolution 105
6.3 Anaphora and coreference processing 109
6.4 Conclusion 112
Chapter 7 Dialogue Acts Recognition 113
7.1 Nature of dialogue acts 114
7.2 Identification and processing of dialogue acts 119
7.3 Multimodal dialogue act processing 122
7.4 Conclusion 124
PART 3 SYSTEM BEHAVIOR AND EVALUATION 125
Chapter 8 A Few Dialogue Strategies 127
8.1 Natural and cooperative aspects of dialogue management 128
8.2 Technical aspects of dialogue management 136
8.3 Conclusion 147
Chapter 9 Multimodal Output Management 149
9.1 Output management methodology 151
9.2 Multimedia presentation pragmatics 156
9.3 Processes 159
9.4 Conclusion 165
Chapter 10 Multimodal Dialogue System Assessment 167
10.1 Dialogue system assessment feasibility 168
10.2 Multimodal system assessment challenges 176
10.3 Methodological elements 180
10.4 Conclusion 190
Conclusion 191
Bibliography 193
Index 203
Über den Autor / die Autorin
Frédéric Landdragin is a computer science engineer and has a PhD from the University of Lorraine, France. He is currently in charge of linguistics research for the French National Center for Scientific Research (CNRS). His studies focus on the analysis and modeling of language interpretation. Man-machine dialogue is one of the applications of this research.
Zusammenfassung
This remarkable work identifies and summarizes the main problems posed by the design of a man-machine dialogue system, then goes on to suggest a number of intriguing ideas that can be used to continue on the path to an efficient, realistic, and fluid communication between human beings and machines.
