EXTENDING FUZZY RELATIONAL DATABASE WITH DISTANCE RELATION AND THE CLUSTERING ALGORITHMS
It is generally considered that the fuzzy database is the next generation of the conventional database be cause the information in the real world is always vague or ambiguous. The theory of fuzzy relational algebraic operations, however, is not so solid as that of conventional relational data model, which has been considered to be the main reason of preventing this theory from widely being appliedo For two fuzzy values, despite of the fuzzy-set-based or the possibility-based type, there does not exist a common way to calculate their closeness or distance measure, even though the closeness relations between single elements on scalar domain have been pre defined in advance. In this paper, we first present a framework of incorporating the possibility-based fuzzy in formation in a conventional relational database structure, as definition 6 and 7, which uniforms the expressing format of determinate values and fuzzy values, and provides a way to handle the determinate value as a special fuzzy value. And then, we extend the often used definition of closeness relation on fuzzy database to distance relation, based on which the most important contribution of this paper, the distance function of fuzzy values, has been presented and proofed, as theorem 1 and Theorem 2. Numeric example has been given to validate the validity. Clustering algorithm on data objects in fuzzy database is therefore promoted.
fuzzy database clustering algorithm distance relation similarity relation
Yiyong Xiao Ikou Kaku Wenbing Chang
Department of System and Engineering, Beihang University, Beijing 10083, China Department of Management Science and Engineering,Faculty of Systems Science and Technology, Akita Pr
国际会议
The Ninth International Conference on Industrial Management(第九届工业管理国际会议 ICIM2008)
日本大阪
英文
694-705
2008-09-16（万方平台首次上网日期，不代表论文的发表时间）