An expert system approach for the choice of appropriate data types in a Fuzzy Database

Article type: Research Article

Authors: Touzi, Amel Grissa | Hassine, Mohamed Ali Ben

Affiliations: National School of Engineers of Tunis, Tunisia | Department of Computer Sciences, Faculty of Sciences of Tunis, Tunisia

Note: [] Corresponding author. E-mail: [email protected] (A.G. Touzi).

Note: [] Corresponding author. E-mail: [email protected] (M.A.B. Hassine).

Abstract: Several applications need to stock and interrogate fuzzy data in a flexible manner. Unfortunately, these two requirements are not supported by the database management systems which currently dominate the databases servers market. Several works have tried to cure these insufficiencies; we speak henceforth of fuzzy databases and flexible querying of these databases. The design problem of such databases did not receive the interest that it deserves. The fuzzy data types are numerous and varied what puts the designer of a fuzzy database in front of a complex problem of choice. In this paper, we propose an approach based on an expert system which helps the designer to choose the appropriate data types for the columns susceptible to stock fuzzy data or interrogated by flexible queries. This solution suits the migration toward a fuzzy database as well as a fuzzy database under design. Although this solution is based on the fuzzy model GEFRED (GEneralised model for Fuzzy RElational Database), it can be applied to other fuzzy models. This expert system can be integrated into the classic design environments like Rational Rose, Power AMC, etc.

Keywords: Fuzzy Databases, Expert Systems, Fuzzy Attributes, FSQL, GEFRED

DOI: 10.3233/IFS-2010-0456

Journal: Journal of Intelligent & Fuzzy Systems, vol. 21, no. 4, pp. 261-273, 2010