Query term weights as constraints in fuzzy information retrieval

• Published in: Information processing & management Vol. 27; no. 1; pp. 15 - 26
• Main Authors: Bordogna, G., Carrara, P., Pasi, G.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 1991; Elsevier Science; Pergamon Press; Elsevier Science Ltd
• Subjects: Boolean; Boolean Logic; Boolean strategies; Evaluation Criteria; Exact sciences and technology; Fuzzy Set Theory; Fuzzy sets; Indexing; Information and communication sciences; Information Retrieval; Information retrieval systems. Information and document management system; Information Science; Information science. Documentation; Information storage and retrieval; Information systems; Information work; Mathematical Models; Online Systems; Queries; Query formulation; Relevance (Information Retrieval); Sciences and techniques of general use; Search Strategies; Searches; Searching; Semantics; Strategies; Studies; Subject indexing; Technical services; User Needs (Information); User Satisfaction (Information); Weighted Term Searching; Boolean expression; Weighting; Information system; Boolean logic; Theory; Information retrieval; Search term; Models; Fuzzy set
• ISSN: 0306-4573; 1873-5371
• DOI: 10.1016/0306-4573(91)90028-K

The generalization of the Boolean information retrieval model has presented several problems, especially with respect to query weighting. Assuming some recent ideas, which outline the need for two levels of query weighting, we propose two different semantics for Boolean expression weights and term weights. By interpreting query term weights as constraints on document representations, we formally define them in the fuzzy set theoretical context and show how this formalization satisfies the wish-list clarified by Cater and Kraft.