Figures p. xi
Foreword p. xv
Preface p. xix
Overview p. 1
Finding Out About--A Cognitive Activity p. 1
Working within the IR Tradition p. 8
Keywords p. 10
Elements of the Query Language p. 10
Topical Scope p. 11
Document Descriptors p. 12
Query Syntax p. 13
Query Sessions p. 14
Documents p. 16
Structured Aspects of Documents p. 19
Corpora p. 20
Document Proxies p. 20
Genre p. 21
Beyond Text p. 22
Indexing p. 26
Automatically Selecting Keywords p. 27
Computer-Assisted Indexing p. 28
FOA versus Database Retrieval p. 29
How Well Are We Doing? p. 34
Summary p. 36
Extracting Lexical Features p. 39
Building Useful Tools p. 39
Interdocument Parsing p. 40
Intradocument Parsing p. 42
Stemming and Other Morphological Processing p. 44
Noise Words p. 47
Summary p. 48
Example Corpora p. 48
Implementation p. 50
Basic Algorithm p. 51
Fine Points p. 54
Software Libraries p. 58
Weighting and Matching against Indices p. 60
Microscopic Semantics and the Statistics of Communication p. 60
Remember Zipf p. 62
Looking for Meaning in All the Wrong Places (At the Character Level) p. 63
Zipf's Own Explanation p. 64
Benoit Mandelbrot's Explanation p. 66
Herbert Simon's Explanation p. 67
More Recent Zipfian Sightings p. 68
Summary p. 71
A Statistical Basis for Keyword Meaning p. 71
Lexical Consequences, Internal/External Perspectives p. 71
Word Occurrence as a Poisson Process p. 73
Resolving Power p. 76
Language Distribution p. 78
Weighting the Index Relation p. 81
Informative Signals versus Noise Words p. 83
Inverse Document Frequency p. 84
Vector Space p. 86
Keyword Discrimination p. 88
Vector Length Normalization p. 89
Summary: SMART Weighting Specification p. 92
Matching Queries against Documents p. 93
Measures of Association p. 94
Cosine Similarity p. 95
Calculating TF-IDF Weighting p. 96
Computing Partial Match Scores p. 97
Summary p. 101
Assessing the Retrieval p. 105
Personal Assessment of Relevance p. 106
Cognitive Assumptions p. 106
Extending the Dialog with RelFbk p. 109
Using RelFbk for Query Refinement p. 111
Using RelFbk to Adapt Documents' Indices p. 115
Summary p. 116
Aggregated Assessment: Search Engine Performance p. 116
Underlying Assumptions p. 116
Consensual Relevance p. 118
Traditional Evaluation Methodologies p. 119
Basic Measures p. 122
Ordering the Retr Set p. 124
Normalized Recall and Precision p. 128
Multiple Retrievals across Varying Queries p. 129
One-Parameter Criteria p. 132
Test Corpora p. 135
Other Measures p. 137
RAVE: A Relevance Assessment VEhicle p. 141
RAVeUnion p. 141
RAVePlan p. 142
Interactive RAVE p. 143
RAVeCompile p. 145
Summary p. 146
Mathematical Foundations p. 149
Derivation of Zipf's Law for Random Texts p. 149
Discussion p. 152
Dimensionality Reduction p. 153
A Simple Example p. 153
Formal Notions of Similarity p. 155
Singular Value Decomposition p. 156
How Many Dimensions k to Reduce to? p. 157
Other Uses of Vector Space p. 158
Computational Considerations p. 159
"Latent Semantic" Claims p. 159
Preference Relations p. 161
Multidimensional Scaling p. 161
Information in RelFbk p. 163
Connections between MDS and LSI p. 165
Clustering p. 165
The Cluster Hypothesis p. 165
Clustering Algorithms p. 166
Probabilistic Retrieval p. 167
Probability Ranking Principle p. 168
Bayesian Inversion p. 169
Odds Calculation p. 170
Binary Independence Model p. 170
Linear Discriminators p. 173
Cost Analysis p. 175
Bayesian Networks p. 175
Inference beyond the Index p. 182
Citation: Interdocument Links p. 185
Bibliometric Analysis of Science p. 187
Time Scale p. 190
Legal Citation p. 191
Citations and Arguments p. 193
Analyzing WWW Adjacency p. 195
Hypertext, Intradocument Links p. 199
Footnotes, Hyperfootnotes, and cf. p. 200
Hierarchic Containment p. 201
Argument Relations p. 205
Intra- versus Interdocument Relations p. 207
Beyond Unary About (k) Predicates p. 209
Keyword Structures p. 210
Automatic Thesaurus Construction p. 211
Corpus-Based Linguistics and WordNet p. 213
Taxonomies p. 217
Social Relations among Authors p. 220
AI Genealogy p. 221
An Empirical Foundation for a Philosophy of Science p. 223
Modes of Inference p. 224
Theorem-Proving Models for Relevance p. 224
Spreading Activation Search p. 225
Discovering Latent Knowledge within a Corpus p. 234
Deep Interfaces p. 238
Geographical Hitlists p. 238
FOA (The Law) p. 242
FOA (Evolution) p. 245
Text-Based Intelligence p. 247
Adaptive Information Retrieval p. 252
Background p. 252
Training against Manual Indices p. 254
Alternative Tasks for Learning p. 255
Sources of Feedback p. 256
Building Hypotheses about Documents p. 258
Feature Selection p. 259
Hypothesis Spaces p. 262
Learning Which Documents to Route p. 263
Widrow-Hoff p. 265
User Drift and Event Tracking p. 266
Classification p. 267
Modeling Documents p. 269
Training a Classifier p. 271
Priors p. 272
Other Approaches to Classification p. 273
Nearest-Neighbor Matching p. 273
Boolean Predicates p. 273
When Irrelevant Attributes Abound p. 274
Combining Classifiers p. 275
Hierarchic Classification p. 277
Information-Seeking Agents p. 279
Exploiting Linkage for Context p. 279
The InfoSpiders Algorithm p. 281
Adapting to "Spatial" Context p. 284
Other Learning Applications and Issues p. 286
Adaptive Lenses p. 286
Adapting to Fluid Language Use p. 288
Symbolic and Subsymbolic Learning p. 288
Conclusions and Future Directions p. 292
Things that Are Changing p. 292
WWW Crawling p. 294
Things that Stay the Same p. 298
The FOA Language Game p. 299
Sperber and Wilson's "Relevance" p. 304
Argument Structures p. 305
User as Portal p. 306
Who Needs to FOA p. 307
Authors p. 308
Scientists p. 309
The Changing Economics of Publishing p. 311
Teachers and Students p. 313
Summary p. 316
(Active) Colophon p. 318
Bibliography p. 321
Index p. 347