Mining Top-K Sequential Patterns in the Data Stream Environment

Sequential pattern mining is a process of extracting useful patterns in data sequences. Existing works on mining Top-K patterns on data streams are mostly for non-sequential patterns. In our framework, we focus on the topic of Top-K sequential pattern mining, where users can obtain adequate amount of interesting patterns. The proposed method can automatically adjust the minimum support during mining each batch in the data stream to obtain candidate patterns. Then candidate patterns are maintained by a tree structure for extracting Top-K sequential patterns. Empirical results show that the proposed method is efficient and scalable.