An Efficient Association Rule Mining Algorithm and Business Application

In this paper, aim at the inefficient problem of the a priori algorithms, we design a new matrix data structure, called cooccurrence matrix, in short COM, to store the data information instead of directly using the transactional database. In COM, any item sets can be randomly accessed and counted without many times full scan of the original transactional database. Based on COM, we first divide association rule into two kinds of rule and then we present an efficient algorithms (COM_mining) to find the valid association rules among the frequent items. Finally we apply COM_mining algorithm and a priori algorithm simultaneously to analyze up-down association relationship between various industry stock blocks of China A stock market. From analytical result we can find that in China A stock market, there are indeed up-down association relationship between various industry stock blocks. At the same time, through comparing COM_mining algorithm and a priori algorithm in this application, we can see, COM_mining is more efficient than a priori.