A change detection method for sequential patterns

a r t i c l e i n f o Recent trends in customer-oriented markets drive many researchers to develop sequential pattern mining algorithms to explore consumer behaviors. However, most of these studies concentrated on how to improve accuracy and efficiency of their methods, and seldom discussed how to detect sequential pattern changes between two time-periods. To help business managers understand the changing behaviors of their customers, a three-phase sequential pattern change detection framework is proposed in this paper. In phase I, two sequential pattern sets are generated respectively from two time-period databases. In phase II, the dissimilarities between all pairs of sequential patterns are evaluated using the proposed sequential pattern matching algorithm. Based on a set of judgment criteria, a sequential pattern is clarified as one of the following three change types: an emerging sequential pattern, an unexpected sequence change, or an added/ perished sequential pattern. In phase III, significant change patterns are returned to managers if the degree of change for a pattern is large enough. A practical transaction database is demonstrated to show how the proposed framework helps managers to analyze their customers and make better marketing strategies. Sequential pattern mining is the technique that explores frequently occurring patterns related to time from a large-scale database. It has been applied to web log mining [5], customer purchase behavior analysis [10], DNA sequence analysis [11], project team coordination [22], retailing management [8], and so on. Previous studies related to sequential pattern mining mainly focused on how to build accurate models or how to discover interesting rules in efficient ways. are some well known algorithms that efficiently identify sequential patterns. In addition, other efforts have concentrated on sequential pattern mining in multi-databases [18] with constraints [7], time-intervals [9], and fuzzy sets [6]. Relatively few attempts have been made to analyze sequential pattern changes in databases collected over time [19]. However, some popular patterns at one time-period may not be valid in another time-period [4]. For example, the sequential pattern " Computer → Memory → Color_Printer " is frequent in the last year. However, this pattern might not be popular in this year but change to " Computer → Memory → Multifunctional_Printer. " If managers cannot capture this dynamic behavior change in time and provide appropriate products or services to their customer, customer attrition will be unavoidable. There have been many existing works focusing on dynamic aspects or comparison …