Data integration constraints for consistent data redundancy in fuzzy databases

Although the problem of data integration in relational databases has been extensively studied, little work has addressed this problem in the context of fuzzy relational databases. Data integration is highly complex in fuzzy relational databases, partially because of the involvement of the resemblance relation. Inconsistent data redundancy may occur when the fuzzy databases to be integrated are associated with different resemblance relation on a given domain. This work presents the notions of consistency constraints, and applies them to the problem of data integration in several fuzzy data models. The constraints ensure that fuzzy databases with different resemblance relations agree to each other regarding data redundancy. In addition a solution for integrating inconsistent fuzzy databases with minimal information loss is provided. © 2008 Wiley Periodicals, Inc.