Survey on Sequential Pattern Mining based on User - Constraints

Abstract:

Sequence data mining, also known as sequential pattern mining, is to find all frequent sub-sequences (called sequential patterns) in a sequence database, the threshold of frequency is specified by the user. In recent years, with the explosion growth of information and big data, this problem trends toward mining with constraints to overcome both effectiveness and efficiency challenges since that the constraints represent for the user’s interest. This paper presents a detailed survey of recent studies on mining sequential pattern and the categories of the constraints. Moreover, it also classifies the main methods and analyzes their advantages and limitations, thereby main approaches and strategies to solve sequential pattern mining problems in the future are presented.