Data modeling with imprecise and linguistic information

Abstract

Data modeling needs to consider imprecise and linguistic information so as to cope with fuzziness and partial knowledge often encountered in decision processes. This paper provides an overview of our efforts on fuzzy extensions to classical data models from two perspectives: one at a conceptual level, and the other at a relational database level, giving rise to fuzzy entity-relationship/enhanced entity-relationship (ER/EER) and relational database (RDB) models respectively. Furthermore, some of the recent developments are discussed, namely λ-specialization in a fuzzy EER model, properties of the extended algebraic operators, and discovery of functional dependencies with degrees of satisfaction.