An Efficient Parallel Approach for Frequent Itemset Mining of Incremental Data

Frequent itemset mining is the essential step of data mining process. Further frequent itemset is a primary data obligatory for association rule mining. The Apriori and FP tree are conventional algorithms for mining frequent itemset and envisaging assoc iation rules based on it for knowledge discovery. The process of updating database continuously is known as incremental data mining. In real life, database updates recurrently where exactly conventional algorithms perform incompetently. If w e could use the previous analys ito incrementally mine the frequent itemset from the updated database, the mining process would become more efficient and cost of mining process would be minimized. In this research, we propose a novel incremental mining scheme w ith a parallel approach for disco vering frequent itemset. It uses a data structure called IMBT. It is a Incremental Mining Binary Tree whic h is used to record the itemset in an efficient way. Furthermore, our approach needs not to predetermine the minimum support threshold and scans the database only once.