Algorithms for Mining Constrained Association Rules
Li Cui · Chinese Journal of Computers · 2000
The problem of discovering association rules has received considerable research attention and some fast algorithms for mining association rules have been developed. The authors consider the problem of discovering constrained association rules between items in a large database of sales transactions. Instead of applying such constraints as a post processing step, integrating them into the mining algorithm can dramatically reduce the execution time. This paper presents two new algorithms called Filtering and Separate for solving this problem that are greatlly different from the known algorithms. The two proposed algorithms can be used separately and also can be used together. By discussing their tradeoffs, it shows that Filtering can decrease the size of concerned database effectively, and Separate outperforms the known algorithms greatlly in the number of candidates generated.