Discusses online search techniques that can be used to identify inaccuracies and inconsistencies in records of online and CD-ROM databases. Methods that can be used to sample a database for errors in descriptor, name, or coded fields are described, including browsing, matching, looking up cross-references, and searching for implausible record…

Discusses artificial intelligence (AI) and considers four potential roles for AI in information retrieval (IR): (1) information characterization, i.e., building a knowledge base; (2) information seeking, i.e., search formulation; (3) support functions, including abstracting; and (4) systems integration. Three examples are given that show how AI…

Discusses searching on the World Wide Web. Highlights include search-engine size; features of the six largest search engines; features of five metasearch engines; intelligent agents; two Web sites for difficult-to-find information, including industry statistics and market data; Internet publishers and databases; specialty search engines; and…

Describes methods for applying natural-language processing for automatic concept-based indexing of full text and methods for exploiting the structure and hierarchy of full-text documents to a large collection of full-text documents drawn from the Health Services/Technology Assessment Text database at the National Library of Medicine. Examines how…

Reviews Web-search tools available to librarians. Discusses the Clever project, (part of the Computer Science Principles and Methodologies Department at the IBM Almaden Research Center, Silicon Valley, California), an algorithm that finds authoritative sources on the Web; Google technology, based on links between Web pages; Ask Jeeves, a…

Describes SWEAR (Systematic Weighting and Ranking), a powers-of-two algorithm that can be used for searching the World Wide Web or any large database that automatically creates discrete, well-defined result sets and displays them in decreasing order of likely relevance. Also discusses fuzzy sets. (Author/LRW)

Compares three Web-based online reference services that are described as a cross between a traditional database service and a Web search engine: Ask Jeeves (http://www.ask.com), The Electric Library (http://www.elibrary.com), and Information Please (http://www.infoplease.com). Concludes that each has strengths and weaknesses and aren't directly…

Provides tips on searching Chemical Abstracts (CA). Discusses coverage, indexing policies, differences between the print and online versions, transliterations, arrangement of the print indexes, abbreviations, and importance of the CA Index Guide for effective searching. (MES)

Describes Features, an intelligent Web search engine that is able to perform real-time adaptive feature (i.e., keyword) and document learning. Explains how Features learns from users' document relevance feedback and automatically extracts and suggests indexing keywords relevant to a search query, and learns from users' keyword relevance feedback…

Presents a prototype system for image retrieval from the Internet using Web mining. Discusses the architecture of the Web image retrieval prototype; document space modeling; user log mining; and image retrieval experiments to evaluate the proposed system. (AEF)

Reports on an experiment that examined whether thesaurus terms, related to query in a specified semantic way (synonyms, narrower terms, related terms, or broader terms) could be identified as having a more positive impact on retrieval effectiveness when added to a query through automatic query expansion. (Contains 54 references.) (Author/LRW)

Describes experiments in the retrieval of spoken documents in multimedia systems. Problems in spoken document retrieval are discussed, and a voice mail application is explained that combines state-of-the-art speech recognition with established document retrieval technologies to provide an effective and efficient retrieval tool. (Author/LRW)

Provides an overview of robots that retrieve World Wide Web documents and index data and then store it in a database. Nine robot-generated databases are described, including record content, services, search features, and sample search results; and sidebars discuss the controversy about Web robots and other resource discovery tools. (LRW)

Describes a tool that allows people to browse several characteristics of federal statistical data contained in 200 Web sites before committing to any specific site. Presents the design process used and results from usability testing for the Relation Browser tool and makes recommendations for comparable applications. (Contains 11 references.)…

Discussion of office document search procedures focuses on variable bit-block compression (VBC) signature files and extends VBC signatures for document ranking. Topics include weighted VBC and aggregate VBC; word size, in bits, of term frequencies; storage and retrieval; logarithmic scaling and linear scaling; and ranking performance. (LRW)