Similarity Retrieval Using Fuzzy Database Systems
Sadaaki Miyamoto · 1995
Similarity retrieval means that data items similar to a given query should be retrieved from a database. The aim of the present paper is to show a method of similarity retrieval using fuzzy propositions as indices of multimedia data and a fuzzy set model for similarity retrieval. The fuzzy set model uses t -norm and a fuzzy thesaurus. Three methods of calculating a similarity index are developed. Moreover appropriate data structures in different database systems including fuzzy relational database and object-oriented database for the similarity retrieval are discussed.