Sequential Pattern Mining

Abstract

Sequential pattern mining deals with data represented as sequences (a sequence contains sorted sets of items). Compared to the association rule problem, a study of such data provides “inter-transaction” analysis (Agrawal & Srikant, 1995). Applications for sequential pattern extraction are numerous and the problem definition has been slightly modified in different ways. Associated to elegant solutions, these problems can match with real-life timestamped data (when association rules fail) and provide useful results.