Application of Rough Sets to Information Retrieval

The aim of the present article is to develop a method of The role of fuzzy sets and rough sets are complemenrough retrieval, namely, an application of the rough set tary or orthogonal in information retrieval. While fuzzy theory to information retrieval. After a brief review of query should be considered in relation to uncertainties in fuzzy sets, rough sets, and a fuzzy logical model for innatural languages, rough sets are related to categorical formation retrieval, rough approximations for retrieved structures of knowledge used in retrieval. For example, data are defined. The approximations are considered for fuzzy thesauri represent relational structures for associatboth crisp and fuzzy cases. A fuzzy set is introduced for the rough boundary, and estimation for the membership ing terms, and rough approximations are concerned with for the results of set operations on the boundary is disclassification structure of them. Thus, fuzzy retrieval uscussed. Rough approximations in cases when hierarchiing a fuzzy thesaurus requests fuzzy association of docucal classes are assumed are considered. Moreover, anments, while a rough retrieval scheme requires retrieval of other approximation by a membership sequence is disdocuments with terms, in the same class of the thesaurus, cussed which refines the foregoing approximations. Illustrative examples are shown. which may be crisp or fuzzy. The use of clustering in information retrieval has been studied by many researchers (e.g., van Rijsbergen, 1979).