Dynamic bit vectors: An efficient approach for mining frequent itemsets

摘要

There are twocommonkindsof data formats to be adopted in data mining. One ishorizontal,andthe other isvertical. Approaches based on vertical data formatshavetheadvantagesofrequiringafewernumber ofdatabasescans and computingitemsetsupportsfast.One of the vertical data representations, bit vector, has recently been widely used formining frequent item sets andhas causedsignificantresults.The sizes of bit vectors for item sets are, however, alwaysthe same,equal tothenumber of transactions inadatabase.In this paper, we propose the scheme of dynamic bit vectors to reduce the memory and the computational time formining frequent item sets from transaction databases.A fast methodfor computing the intersectionoftwodynamic bit vectors and an algorithm for miningfrequent item setsbased on the scheme arepresented.The proposed algorithmis also compared with some other approaches and experimental results showthat it isquiteefficientin both the mining time and the memory usage.