Incorporation of Data Mining and Domain Knowledge for Discovering Quasi-Equivalence Database Relationships

Abstract

In this paper, we propose to incorporate domain knowl- edge in a generalized aÆnity-based association rule min- ing algorithm to reduce the size of the data so that only the potentially interesting and relevant portion of the data will be used in the computation procedure. The association rule mining algorithm is an approach to discover the quasi-equivalence relationships of the me- dia objects for databases. After the incorporation of domain knowledge, the computational performance of the mining algorithm can be improved. Experimental results are provided and analyzed.