Analyzing inference in fuzzy database systems

摘要

This paper deals with inference in fuzzy relational databases. It focuses on functional dependencies (FDs) and defines the important notion of a fuzzy inference channel. The analysis is performed using a special abstract model which captures and articulates subjective equivalences underlying fuzziness. Core database integrity properties are defined for the abstract model. They preserve a vital link to the classical model and to existing fuzzy relational database model. The link increases the utility of the inference formalism in applications such as database security and knowledge discovery; these applications are discussed in the context of fuzzy relational databases.