Mining frequent patterns based on IS/sup +/-tree
Haibing Ma, Jin Zhang, Ying-Jie Fan, Yunfa Hu · 2005
Frequent patterns mining play an important role in data mining research. It is the groundwork of other data mining tasks. A novel algorithm is presented for mining frequent patterns based on static IS/sup +/-tree, and is compared extensively with other classical algorithms such as Apriori and FP-growth. The algorithm builds frequent patterns directly, instead of using high-cost candidate sets generation-and-test method adopted by Apriori; it works on a static IS/sup +/-tree, instead of costly dynamic trees adopted by FP-growth; it consumes smaller size of main memory and is more efficient than others. Above all, IS-tree is a general index model and has been widely used in full text storage and index, time series patterns mining and many other fields.