상세정보

Thinking Between the Lines targets a challenge at the heart of the artificialintelligence enterprise: the design of programs that can read and reason on the basis of writtencausal descriptions such as those that appear in encyclopedias, user manuals, and related sources.This capability of "thinking between the lines" -- codified in terms of a task called "causalreconstruction" -- bears directly on the larger question of how computers can usefully exploit thevast repertory of human knowledge concerning causal phenomena.Central to the approach presented is acognitively inspired representation called "transition space," implemented in a program calledPATHFINDER. The transition space representation embodies a conceptual shift from viewing the worldprimarily in terms of states -- or instantaneous snapshots of activity -- to viewing it primarily interms of transitions -- ensembles of changes that can be articulated in language. Transitions,according to this view, serve as antecedents and consequents of causality, and the space of allpossible transitions -- or transition space -- serves as an arena for working out paths ofassociation between the events mentioned within particular causal descriptions.Thinking between theLines provides a computational framework and approach for realizing the significant opportunitiesthat arise for intelligent, automated handling of technical material -- in routing information,answering questions, elaborating or summarizing information to meet the needs of particularindividuals, and performing other useful tasks.Artificial Intelligence series

정보제공 :

CONTENTS
Series Foreword = ⅸ
Acknowledgments = xi
1 Introduction = 1
 1.1 What This Book Is About = 1
 1.2 The Causal Reconstruction Task = 5
 1.3 The Transition Space Representation = 9
 1.4 The PATHFINDER Program = 22
 1.5 A Note to the Reader = 33
2 Causal Reconstruction = 35
 2.1 A Closer Look at the Problem = 35
 2.2 Task Restrictions for PATHFINDER = 39
3 Transition space = 49
 3.1 Guidelines from Perceptrual Psychology = 49
 3.2 Representing Transitions and Events = 52
 3.3 Using Language to Generate Representations = 62
4 Matching in Transition Space = 69
 4.1 Overview = 69
 4.2 Direct Matches between Referenced Events = 73
 4.3 Matches Involving Precedent Events = 91
5 Inference, Background Statements and Assumptions = 99
 5.1 Overview = 99
 5.2 Employing Inference = 106
 5.3 Making Use of Background Statements = 117
 5.4 Identifying Supporting Assumptions = 123
6 Exploratory Transformations = 127
 6.1 Overview = 127
 6.2 Information-Preserving Transformations = 136
 6.3 Non-Information-Preserving Transformations = 147
7 Making Use of Connecting statements = 169
 7.1 Overview = 169
 7.2 Temporal Ordering Statements = 174
 7.3 Other Specifications of Association = 178
8 An Extended Example = 185
 8.1 Phase 1 : Parsing and Encoding the Input = 185
 8.2 Phase 2 : Applying Exploratory Transformations = 196
 8.3 Phase 3 : Associating the Events = 198
 8.4 Phase 4 : Answering Questions = 204
9 Related Literature = 213
 9.1 Research in Artificial Intelligence = 213
 9.2 Research in Psychology = 224
 9.3 Research in Linguistics and Philosophy = 228
10 Conclusions = 233
 10.1 Contributions of the Research = 233
 10.2 Extending the Approach = 237
 10.3 New Horizons = 241
A PATHFINDER Implementation = 251
B PATHFINDER Test Examples = 259
Glossary = 275
Bibliography = 283
Index = 295