Finding Out About

A Cognitive Perpective on Search Engine Technology and the WWW

Richard K. Belew

356 pages, parution le 27/09/2004

Ajouter à une liste

Indisponible

Résumé

The World Wide Web is rapidly filling with more text than anyone could have imagined even a short time ago, but the task of isolating relevant parts of this vast information has become just that much more daunting. Richard Belew brings a cognitive perspective to the study of information retrieval as a discipline within computer science. He introduces the idea of Finding Out About (FQA) as the process of actively seeking out information relevant to a topic of interest and describes its many facets - ranging from creating a good characterization of what the user seeks, to what documents actually mean, to methods of inferring semantic dues about each document, to the problem of evaluating whether our search engines are performing as we have intended.

Finding Out About explains how to build the tools that are useful for searching collections of text and other media. In the process it takes a close look at the properties of textual documents that do not become clear until very large collections of them are brought together and shows that the construction of effective search engines requires knowledge of the statistical and mathematical properties of linguistic phenomena, as well as an appreciation for the cognitive foundation we bring to the task as language users. The unique approach of this book is its even handling of the phenomena of both numbers and words, making it accessible to a wide audience.

The textbook is usable in both undergraduate and graduate classes on information retrieval, library science, and computational linguistics. The text is accompanied by a CD-ROM that contains a hypertext version of the book, including additional topics and notes not present in the printed edition. In addition, the CD contains the full text of C. J."Keith" van Rijsbergen's famous textbook, Information Retrieval(now out of print). Many active links from Belew's to van Rijsbergen's hypertexts help to unite the materiel. Several test corpora and indexing tools are provided, to support the design of your own search engine. Additional exercises using these corpora and code are available to instructors. Also supporting this book is a Web site that will include recent additions to the book, as well as links to sites of new topics and methods.

Sommaire

Overview
Extracting lexical features
Weighting and matching against indices
Assessing the retrieval
Mathematical foundations
Inference beyond the index
Adaptive information retrieval
Conclusions and future directions

Voir tout

Replier

Caractéristiques techniques

	PAPIER
Éditeur(s)	Cambridge University Press
Auteur(s)	Richard K. Belew
Parution	27/09/2004
Nb. de pages	356
Format	18 x 24
Couverture	Relié
Poids	788g
Intérieur	Noir et Blanc
EAN13	9780521630283
ISBN13	978-0-521-63028-3