Fuzzy Associative Summaries for Databases

The paper introduces the problem of mining Fuzzy Associative Summaries. An example of summary is (Age; Medium) ! (Salary; Low) ! (Department; A); 0:34; it means that when the Age is Medium, this implies that the Salary is Low, and this further implies that the Department is A, with a support of 0:34. In this paper, we define the notion of fuzzy associative summary, and propose an algorithm to mine them from within a relational table. Then, we discuss our implementation on top of Soft-SQL, the extension of the classical SQL query language to express flexible queries based on soft operators and sets of linguistic predicates on classical relational databases [1], and present some experimental results.