Interpretation of the Keywords Weights in Information Retrieval: Fuzzy Logic Based Approaches

Logical models offer a sound theoretical framework for information retrieval (IR). Fuzzy logic provides a convenient extension of classic logic based approaches. In particular it makes it possible to grasp the graduality of some relevant concepts and to model both imprecision and uncertainty inherent to the retrieval process. In the paper we show a few possible fuzzy logic based approaches to information retrieval. In particular we study how well-known keywords weights semantics may be recovered in some more advanced models of this type.