Efficient Strategies for Average Constraint-Based Sequential Pattern Mining

Sequential pattern mining based on constraint is now an important research direction of data mining, since it can reduce the generation of useless candidates as well as make the generated patterns meet the requirements of special users. Average value constraint is a kind of tough aggregate constraint. We propose here an effective pruning strategy based on average value constraint to avoid constructing unnecessary projected database and a novel frequent sequential pattern mining algorithm incorporating above strategy. An algorithm called SMAC (sequential frequent pattern mining with average constraints) was proposed and designed here based on Prefix Span method . At last, the algorithm was analyzed by experiment to show that the proposed method is more effective than Prefix Growth.