WSpCPs: Weighted sequential pattern mining based on cluster-pruning mechanism

One of the major important problems in sequential pattern mining is the explosion of the number of results. To solve this problem, a new algorithm, called weighted sequential pattern mining based on cluster-pruning strategy (WSpCPs), is proposed in this paper. The purpose of our algorithm is to select some high-quality sequences that describe the full result. It utilizes I-step and S-step operations to generate new sequences and their bitmaps in iterative process. WSpCPs proposes cluster-pruning strategy to select sequences from the full result. Experiments show that WSpCPs is an efficient method to reduce the number of result.