Improved algorithm based on Sequential pattern mining of big data set

Abstract

With the Sequential pattern mining is one of the most important research in data mining.t. Comparing with other data mining algorithms,sequential pattern mining algorithm is mainly based on ordered datasets to mine sequential patterns, it has the advantages of high frequency, practicality. Therefore, sequential pattern mining has been widespread concerned and in-depth researched by experts and scholars at home and abroad. So far sequential pattern mining has a lot of classic algorithms. Among those algorithms,Prefix Span algorithm is one of the most widely used algorithms. The algorithm uses the prefix projection technology, which could avoid candidate items effectively, thus to improve the efficiency of mining in a certain extent. However, Prefix Span algorithm needs to structure a lot of projection database, besides, structuring projection database not only need consume huge memory, but increase a lot of scan time. Therefore, this paper improves Prefix Span algorithm and proposes BLSPM algorithm. This algorithm can greatly reduce the numbers of construction projection database, thus improving the efficiency of sequential pattern mining.