Information Retrieval -- Contents -- Foreword -- Preface -- About the Editors -- List of Contributors -- Introduction -- 1 Information Retrieval Models -- 1.1 Introduction -- 1.1.1 Terminology -- 1.1.2 What is a model? -- 1.1.3 Outline -- 1.2 Exact Match Models -- 1.2.1 The Boolean model -- 1.2.2 Region models -- 1.2.3 Discussion -- 1.3 Vector Space Approaches -- 1.3.1 The vector space model -- 1.3.2 Positioning the query in vector space -- 1.3.3 Term weighting and other caveats -- 1.4 Probabilistic Approaches -- 1.4.1 The probabilistic indexing model -- 1.4.2 The probabilistic retrieval model -- 1.4.3 The 2-Poisson model -- 1.4.4 Bayesian network models -- 1.4.5 Language models -- 1.4.6 Google's PageRank model -- 1.5 Summary and Further Reading -- Exercises -- References -- 2 User-centred Evaluation of Information Retrieval Systems -- 2.1 Introduction -- 2.1.1 Background -- 2.1.2 Chapter outline -- 2.2 The MEDLARS Test -- 2.2.1 Description of Lancaster's test of MEDLARS -- 2.2.2 Evaluation characteristics -- 2.3 The Okapi Project -- 2.3.1 The objectives of Okapi -- 2.3.2 Okapi at TREC -- 2.3.3 The impact of Okapi -- 2.4 The Interactive IR Evaluation Model -- 2.4.1 The cognitive IR approach -- 2.4.2 The three parts of the IIR evaluation model -- 2.5 Summary -- Exercises -- References -- 3 Multimedia Resource Discovery -- 3.1 Introduction -- 3.2 Basic Multimedia Search Technologies -- 3.2.1 Piggy-back text retrieval -- 3.2.2 Automated annotation -- 3.2.3 Content-based retrieval -- 3.2.4 Fingerprinting -- 3.3 Challenges of Automated Visual Indexing -- 3.4 Added Services -- 3.4.1 Video summaries -- 3.4.2 New paradigms in information visualisation -- 3.4.3 Visual search and relevance feedback -- 3.5 Browsing: Lateral and Geotemporal -- 3.6 Summary -- Exercises -- References -- 4 Image Users' Needs and Searching Behaviour -- 4.1 Introduction.

4.2 Image Attributes and Users' Needs -- 4.2.1 Image attributes -- 4.2.2 Image attributes in queries -- 4.2.3 Attributes beyond queries -- 4.2.4 Image needs -- 4.3 Image Searching Behaviour -- 4.3.1 Search process -- 4.3.2 Search strategies -- 4.3.3 Relevance criteria -- 4.4 New Directions for Image Access -- 4.4.1 Social tagging -- 4.4.2 Images in context -- 4.4.3 Visualisations -- 4.4.4 Workspaces -- 4.5 Summary -- Exercise -- References -- 5 Web Information Retrieval -- 5.1 Introduction -- 5.2 Distinctive Characteristics of the Web -- 5.2.1 Web data -- 5.2.2 Web structure -- 5.2.3 User behaviour -- 5.2.4 User interaction data -- 5.3 Three Ranking Problems -- 5.3.1 Retrieval -- 5.3.2 Selective crawling -- 5.3.3 Index organisation -- 5.4 Other Web IR Issues -- 5.4.1 Stemming -- 5.4.2 Treatment of near-duplicate content -- 5.4.3 Spelling suggestions -- 5.4.4 Spam rejection -- 5.4.5 Adult content filtering - genre classification -- 5.4.6 Query-targeted advertisement generation -- 5.4.7 Snippet generation -- 5.4.8 Context and web information retrieval -- 5.5 Evaluation of Web Search Effectiveness -- 5.5.1 TREC-9 Web Track: Realistic queries, rich link structure, traditional IR task -- 5.5.2 Evaluation using web-specific tasks -- 5.5.3 Future directions for web IR evaluations -- 5.5.4 Comparing results lists in context -- 5.5.5 Evaluation by commercial web search companies -- 5.6 Summary -- Exercises -- References -- 6 Mobile Search -- 6.1 Introduction: Mobile Search - Why Now? -- 6.1.1 Technological drivers -- 6.1.2 Predicted demand for mobile search -- 6.2 Information for Mobile Search -- 6.2.1 Linking information to physical space -- 6.2.2 The storage of information -- 6.2.3 The ownership of information -- 6.3 Designing for Mobile Search -- 6.3.1 Characteristics of mobile usage -- 6.3.2 Filters as a framework for mobile search.

6.3.3 Manual versus automatic filtering -- 6.3.4 Using filters to push information -- 6.4 Case Studies -- 6.4.1 Oslo airport - AmbieSense -- 6.4.2 The Swiss Alps - WebPark -- 6.5 Summary -- Exercises -- References -- 7 Context and Information Retrieval -- 7.1 Introduction -- 7.2 What is Context? -- 7.2.1 Whose context? -- 7.3 Context in Information Retrieval -- 7.3.1 Context in the wider sense -- 7.3.2 Perceptions of context in related fields -- 7.3.3 Example: context and images -- 7.4 Context Modelling and Representation -- 7.4.1 Context modelling -- 7.4.2 User models and their relationship to context -- 7.4.3 Past, present and future contexts -- 7.5 Context and Content -- 7.5.1 Representation of context -- 7.5.2 Capturing context -- 7.5.3 Searching with context information -- 7.5.4 Context templates -- 7.6 Related Topics -- 7.6.1 Personalisation and context -- 7.6.2 Mobility and context -- 7.7 Evaluating Context-aware IR Systems -- 7.7.1 Principles of methodology -- 7.8 Summary -- Exercises -- References -- 8 Text Categorisation and Genre in Information Retrieval -- 8.1 Introduction: What is Text Categorisation? -- 8.1.1 Purpose of categorisation -- 8.2 How to Build a Text Categorisation System -- 8.2.1 The classifier component -- 8.2.2 The machine learning component -- 8.2.3 The feature selection component -- 8.3 Evaluating Text Categorisation Systems -- 8.4 Genre: Text Structure and Purpose -- 8.4.1 An overview of genre -- 8.4.2 Text categorisation and genre -- 8.4.3 The importance of layout -- 8.5 Related Techniques: Information Filtering -- 8.6 Applications of Text Categorisation -- 8.7 Summary and the Future of Text Categorisation -- Exercises -- References -- 9 Semantic Search -- 9.1 Introduction -- 9.1.1 Limitations of current search technology -- 9.1.2 Ontologies -- 9.1.3 Knowledge bases and semantic repositories -- 9.2 Semantic Web.

9.2.1 Semantic web and semantic search -- 9.2.2 Basic semantic web standards: RDF(S) and OWL -- 9.3 Metadata and Annotations -- 9.4 Semantic Annotations: the Fibres of the Semantic Web -- 9.5 Semantic Annotation of Named Entities -- 9.5.1 Named entities -- 9.5.2 Semantic annotation model and representation -- 9.6 Semantic Indexing and Retrieval -- 9.6.1 Indexing with respect to lexical concepts -- 9.6.2 Indexing with respect to named entities -- 9.6.3 Retrieval as spreading activation over semantic network -- 9.7 Semantic Search Tools -- 9.7.1 Searching through document-level RDF annotations - QuizRDF -- 9.7.2 Exploiting massive background knowledge - TAP -- 9.7.3 Character-level annotations and massive world knowledge - KIM -- 9.7.4 Squirrel -- 9.7.5 Other approaches -- 9.8 Summary -- Exercises -- References -- 10 The Role of Natural Language Processing in Information Retrieval: Searching for Meaning and Structure -- 10.1 Introduction -- 10.2 Natural Language Processing Techniques -- 10.2.1 Named entity recognition -- 10.2.2 Information extraction -- 10.2.3 WordNet -- 10.2.4 Word sense disambiguation -- 10.2.5 Evaluation -- 10.3 Applications of Natural Language Processing in Information Retrieval -- 10.3.1 Text mining -- 10.3.2 Question answering -- 10.4 Discussion -- 10.5 Summary -- Exercises -- References -- 11 Cross-Language Information Retrieval -- 11.1 Introduction -- 11.2 Major Approaches and Challenges in CLIR -- 11.3 Identifying Translation Units -- 11.3.1 Tokenisation -- 11.3.2 Stemming -- 11.3.3 Phrase identification -- 11.3.4 Stop-words -- 11.4 Obtaining Translation Knowledge -- 11.4.1 Obtaining bilingual dictionaries and corpora -- 11.4.2 Extracting translation knowledge -- 11.4.3 Dealing with out-of-vocabulary terms -- 11.5 Using Translation Knowledge -- 11.5.1 Translation disambiguation -- 11.5.2 Weighting translation alternatives.

11.5.3 Using translation probabilities in term weighting -- 11.6 Interactivity in CLIR -- 11.6.1 Interactive CLIR -- 11.6.2 Query translation in interactive CLIR -- 11.6.3 Document selection in interactive CLIR -- 11.7 Evaluation of CLIR Systems -- 11.7.1 Cranfield-based evaluation framework -- 11.7.2 Evaluations on interactive CLIR -- 11.7.3 Current CLIR evaluation frameworks -- 11.8 Summary and Future Directions -- 11.8.1 Current achievements in CLIR -- 11.8.2 Future directions for CLIR -- 11.8.3 Further reading -- Exercises -- References -- 12 Performance Issues in Parallel Computing for Information Retrieval -- 12.1 Introduction -- 12.2 Why Parallel IR? -- 12.3 Review of Previous Work -- 12.4 Distribution Methods for Inverted File Data -- 12.4.1 On-the-fly distribution -- 12.4.2 Inverted file replication -- 12.4.3 Inverted file partitioning -- 12.5 Tasks in Information Retrieval -- 12.5.1 The indexing task -- 12.5.2 The probabilistic search task -- 12.5.3 The passage retrieval task -- 12.5.4 The routing/filtering task -- 12.5.5 The index update task -- 12.6 A Synthetic Model of Performance for Parallel Information Retrieval -- 12.7 Empirical Examination of Synthetic Model -- 12.7.1 Comparative results using indexing models -- 12.7.2 Comparative results using search models -- 12.7.3 Comparative results using passage retrieval models -- 12.7.4 Comparative results using term selection models -- 12.7.5 Comparative results using index update model -- 12.8 Summary and Further Research -- Exercises -- References -- Solutions to Exercises -- Index.