数据流上的频繁闭项集挖掘算法-计算机工程.PDF

36 18 2010 9 Vol.36 No.18 Computer Engineering September 2010 ·· 2010 A TP311 ( 410073) MFWSR MFWSR Algorithm of Mining Frequent Closed Itemsets on Data Streams TAO Ke, WANG Yi-jie (National Key Laboratory for Parallel and Distributed Processing, School of Computer, National University of Defense Technology, Changsha 410073, China) AbstractThis paper proposes an algorithm of Mining Frequent closed itemsets with Window Sliding Rapidly(MFWSR) against the complexity of data structure and process for determination. With the data stream represented by compact data structure, process for determination simplified, MFWSR improves the temporal and spatial efficiency while it can response to the requests of user-specified support threshold. Experimental results show that compared with existing algorithms, MFWSR achieves higher temporary and spatial efficiency while the accuracy remains. Key wordsdata stream; data mining; frequent closed itemsets 1 1 W (k 2k Wi,j i j sup (X ) X ) str (X ) X len(X ) X D T T i i 2.2 ()k 1 D T k f (T ) {i ∈I | ∀∈t T , i ∈T} 2 D X [1] g