Mining Self-Adaptive Sequence Patterns Based on the Sequence Fuzzy Concept Lattice

Traditional algorithms in mining sequence patterns only discover the frequent sequences satisfying the minimum support threshold minsup; however, these methods don´t consider the importance of actual sequences. In order to mine the important sequence patterns satisfying users´ many demands, this paper presents a new algorithm called SASeqP for mining self-adaptive sequence patterns based on sequence fuzzy concept lattice. The algorithm firstly introduces the weight for each item in sequences. On the basis of the weight, we define the sequence fuzzy formal context and sequence fuzzy concept lattice. After these, we define self-adaptive sequence patterns and the self-adaptive coefficient that may dynamically adjust the minimum support threshold minsup. At last, this article presents the new algorithm SASeqP for mining self-adaptive sequence patterns. The experimental results show that the algorithm SASeqP has excellent performance on the time-spatial complexity.