An Effective High Utility Itemset Mining Algorithm with Big Data Based on MapReduce Framework

摘要

Incremental data mining is an important factor of data mining. Through incremental data mining the results are improved and becomes more reliable. It brings efficiency in the mining process. The data grows exponentially as the time passes and difficulties in handling the data also increases. Incremental data mining helps to find high profit and high utility pattern mining. Frequent Itemset Mining(FIM) is widely used but it has some important limitations when it comes to analyse the customer transactions. One important limitation is that purchase quantities are not taken into account. Thus an item can appear only once or zero time in transaction. Second limitation of FIM is that all items have same importance in transactions. Hence, frequent pattern generated by FIM may not be highly profitable. In this paper we propose algorithm with big data to find high utility itemsets. The algorithm uses map-reduce framework and parallel processing to find high profit frequent mining.