A Novel Algorithm for Association Rule Mining without Candidate

A-priori is an influence algorithm for finding frequent item sets from association rules. But there are two hard questions may be involved for average users during finding frequent candidates. One question is massive amounts of candidates and the other is that set support count threshold for every level candidate generations. This paper discusses one algorithm called And, which is usually used in logical algorithms, And Code (AC) algorithm can discover frequent itemsets without producing candidates and setting thresholds for candidates. Frequent itemsets can be fast discovered by corresponding codes which are cited by this paper to describe different itemsets for AC algorithm. The support count of frequent itemsets can be computed before the process of scanning during processes of AC algorithm. The codes of itemsets are defined and detailed in section two. Lastly, an example is presented to detail processes of AC algorithm and test that AC algorithm can more efficiently find frequent without candidates than a-priori algorithm which may produce a large of candidates during scan process.