Efficient Mining of Frequent Episodes from Complex Sequences

摘要

Discovering patterns with highly significance is an important problem in data mining discipline. Anepisode is defined to be a partially ordered set of events for a consecutive and fixed time intervals in asequence. Previous studies in episodes consider only frequent episodes in a sequence of events (calledsimple sequence). In real world, we may find a set of events at each time slot in terms of various intervals(hours, days, weeks, etc.) We refer to such sequences as complex sequences. Mining frequent episodesin complex sequences has more extensive applications than in simple sequences. In this paper, wediscuss the problem on mining frequent episodes in a complex sequence. We extend previous algorithmMINEPI to MINEPI+ for episode mining from complex sequences. Furthermore, a memory-anchoredalgorithm called EMMA is introduced for the mining task. Experimental evaluation on both real worldand synthetic data sets shows that EMMA is more efficient than MINEPI+.