Figures
Finding Out About
1.1 The FOA Conversation Loop1.2 Retrieval of Documents in Response to a Query
1.3 Assessment of the Retrieval
1.4 Schematic of Search Engine
1.5 A Query Session
1.6 Finding Out About POLITICAL FIDELITY
1.7 Obsolete Concert Schedule
1.8 Results of SCSI Search of UseNet
1.9 A Relevant Posting
1.10 Comparison of Retrieved versus Relevant Documents
2.1 Parsing Email and AIT to Common Specifications
2.2 Finite State Machine
2.3 AIT Year Distribution
2.4 Basic Postings Data Structures
2.5 Refined Postings Data Structures
2.6 STAIRS Posting Information
2.7 Quoted Lines in an Email Message
3.1 Zipfian Distribution of AIT Words
3.2 Rank/Frequency Distribution of Click-Paths
3.3 Resolving Power
3.4 Specificity/Exhaustivity Trade-Offs
3.5 Indexing Graph
3.6 Hypothetical Word Distributions
3.7 Vector Space
3.8 Length Normalization of Vector Space
3.9 Sensitivity of IDF to “Document” Size
3.10 Pivot-Based Document Length Normalization
4.1 Relevance Scale
4.2 RelFbk Labeling of the Retr Set
4.3 Query Session, Linked by RelFbk
4.4 RelFbk Labels in Vector Space
4.5 Various Ways of Being Irrelevant
4.6 Using RelFbk to Refine the Query
4.7 Document Modifications due to RelFbk
4.8 Consensual Relevance
4.9 Relevant versus Retrieved Sets
4.10 Recall/Precision Curve
4.11 Instability of Beginning of Re/Pre Curve
4.12 Best/Worst Retrieval Envelope
4.13 Normalized Recall
4.14 Multiple Queries, Fixed Recall Levels
4.15 11-Point Average Re/Pre Curves
4.16 TREC Query
4.17 Distinguishing between Overlapping Distributions
4.18 Operating Characteristic Curve
4.19 RAVE Interface
5.1 Lexicographic Tree Underlying Zipfian Distribution
5.2 a as Function of M, Number of Distinct Characters
5.3 Weight and Height Data Reduction
5.4 SVD Decomposition
5.5 Retrieval Performance as Function of SVD Dimension (k)
5.6 Random variables underlying Binary Independence Model
5.7 Interaction between Parental Influences
5.8 Bayesian Network Representation for FOA
5.9 Concept-Matching Version of Bayesian Network
5.10 Two Examples of Query Modeling
6.1 Other Information Available for FOA
6.2 Basic Structure of Citations
6.3 Temporal Structure in Citations
6.4 Rose’s Two-Dimensional Analysis of Shepherd Treatment Codes
6.5 Citation-Expanded Hitlist
6.6 Containment and Document References
6.7 Correlation of Passages
6.8 Visualizing Topical Distributions
6.9 Topical Document Distributions
6.10 Intradocument Relations
6.11 FOA Overview
6.12 Typography Used to Isolate Talmudic Voices
6.13 HTML Version of the Talmud
6.14 Example of the Term LYMPHOMA within the MeSH Keyword Thesaurus
6.15 Bipolar Organization of Adjectives in WordNet
6.16 Nesting Taxonomies from Various Sources
6.17 A Sample of the AI Genealogical Record
6.18 Spreading Activation Search
6.19 Subnet Corresponding to Each Document
6.20 AIR Interface
6.21 Semantic Net of Legal Document Relations
6.22 Swanson’s Search for Latent Knowledge
6.23 CIVIL WAR BATTLE Query, Standard Textual Hitlist
6.24 CIVIL WAR BATTLE Query, Geographical Presentation
6.25 The Long Life of Legal Documents
6.26 The Annotation Relation between Text and Sequence Data
7.1 Learning Conceptual Structures
7.2 Browsing across Queries in Same Session
7.3 Inductive Bias
7.4 Document Modifications due to RelFbk
7.5 Training a Classifier
7.6 RIPPER Classification Rule
7.7 Covering Algorithms
7.8 Combining Experts
7.9 Optimal Weightings Distribution
7.10 Hierarchic Classification
7.11 Ancestors of a Class
7.12 Adaptive Lens
8.1 Crawler Coverage
8.2 A Semiotic View of FOA
8.3 A Semiotic Analysis of Keyword Mismatch
8.4 Grice’s Maxims
8.5 Query as Portal, Connecting Corpora
8.6 The Tell/Ask Duality