Frequent itemset mining on hadoop

摘要

One of the most important problems in data mining is frequent itemset mining. It requires very large computation and I/O traffic capacity. For that reason several parallel and distributed mining algorithms were developed. Recently the mapreduce distributed data processing paradigm is unavoidable and porting the current algorithms to mapreduce is in focus. In this paper a substantial frequent itemset mining algorithms and their mapreduce implementations are introduced and investi-gated. An algorithm improvement is also proposed and analyzed.