No label defined (Q4406071)

The book concerns a traditional discipline called information retrieval (IR). In contrast to previous monographs about IR, the author extends this discipline to modern approaches including, e.g., fuzzy logic and data mining methods. Last but not least, interactions of IR and web searching are presented. Chapters 1 and 2 explain fundamentals of classic IR methods to the reader. Chapter 3 contains traditional text models like Boolean and vector space models. Other techniques like classifications, thesauri, and semantic networks are mentioned. For evaluating and comparing particular models, basic notions like relevance and its measuring by precision and recall coefficients are included. The author emphasises the role of various standard test collections (benchmarks), mainly TREC, employed in IR. In the second part of the book the reader finds basics of knowledge discovery procedures and data mining in the context of information retrieval. Chapter 4 shortly describes learning as a data processing method. Various categorization methods are discussed in Chapter 5 including algorithms like ID3. Also other well-known methods such as rule systems, decision trees are summarized here. Chapter 6 is devoted to clustering and fuzzy set approaches. Associative rules are presented in Chapter 7. The described methods are explicated in a complex example in Chapter 8. The third part of the book contains extended retrieval approaches. The vector model considered as a fuzzy set approach is discussed in Chapter 9. Particularly, generalized Boolean methods are presented here. Chapter 10 introduces a probabilistic approach to IR. IR models based on logic are mentioned in Chapter 11 in the context of knowledge representation. Applications and experiments with TREC collections are discussed in Chapter 12. A well-known method, latent semantic indexing, appropriate for dimension reduction in vector-oriented models is discussed in Chapter 13 on corpus-oriented approaches. The last part of the book shortly mentions important material about association between IR and Web searching. Chapter 14 describes mark-up languages such as XML and HTML and basics of the recent query languages XPath and XQuery. Chapter 15 continues this topic by a discussion of metadata. Attention is devoted to the standards Dublin Core, PICS, RDF, and to the so-called Semantic Web. Finally, Chapter 16 discusses searching the Web in terms of IR and continues with current specific crawler techniques and ranking algorithms like PageRank. Also, distributed collections and some issues of digital libraries are considered. The material is presented on a sufficiently formal level with a lot of illustrating examples and a rich list of literature items. The book is valuable both for students and for researchers starting in the given area.