Mining of condensed sequential pattern bases

Abstract

Conventional sequential pattern mining methods may meet inherent difficulties when a sequence database is large and/or when sequential patterns to be mined are numerous and/or long, since the number of frequent sequential patterns generated is often too large. In many applications it is sufficient to generate only frequent sequential patterns with support frequency in close-enough approximation instead of in full precision. In this paper, we introduce the concept of condensed frequent sequential pattern-base with guaranteed maximal error bound and develop an algorithm to mine such a condensed sequential pattern-base. Our results show that computing condensed frequent sequential pattern base is promising.