Sequential pattern mining using PrefixSpan with pseudoprojection and Separator Database

Abstract

Sequential pattern mining is a new branch of data mining science that solves inter-transaction pattern mining problems. A comprehensive performance study has been reported that PrefixSpan, one of its algorithms, outperforms GSP, SPADE, as well as FreeSpan in most cases, and PrefixSpan integrated with pseudoprojection technique is the fastest among those tested algorithms. Nevertheless, Pseudoprojection technique, which requires maintaining and visiting the in-memory sequence database frequently until all patterns are found, consumes a considerable amount of memory and induces the algorithm to undertake redundant and unnecessary checks to this copy of original database into memory when the candidate patterns are examined. In this paper, we propose Separator Database to improve PrefixSpan with pseudoprojection through early removal of uneconomical in-memory sequence database. The experimental results show that Separator Database improves PrefixSpan with pseudoprojection. Future research includes exploring the use of Separator Database in PrefixSpan with pseudoprojection to improve mining constrained sequential patterns.