Mining Partially-Ordered Episode Rules in an Event Sequence

Episode rule mining is a popular data mining task for analyzing a sequence of events or symbols. It consists of identifying subsequences of events that frequently appear in a sequence and then to combine them to obtain episode rules that reveal strong relationships between events. But a key problem is that each rule requires a strict ordering of events. As a result, similar rules are treated differently, though they in practice often describe a same situation. To find a smaller set of rules that are more general and can replace numerous episode rules, this paper introduces a novel type of rules called partially-ordered episode rules, where events in a rule are partially ordered. To efficiently find all these rules in a sequence, an efficient algorithm named POERM (PartiallyOrdered Episode Rule Miner) is presented. An experimental evaluation on several benchmark dataset shows that POERM has excellent performance.

[1]  Jin Wang,et al.  Large-Scale Frequent Episode Mining from Complex Event Sequences with Hierarchies , 2019, ACM Trans. Intell. Syst. Technol..

[2]  Yun Sing Koh.  DRAFT 0 Unsupervised Rare Pattern Mining : A Survey , 2016 .

[3]  Philippe Fournier-Viger,et al.  Mining Episode Rules from Event Sequences Under Non-overlapping Frequency , 2021, IEA/AIE.

[4]  Longbing Cao,et al.  Mining Partially-Ordered Sequential Rules Common to Multiple Sequences , 2015, IEEE Trans. Knowl. Data Eng..

[5]  Heikki Mannila,et al.  Discovering Frequent Episodes in Sequences , 1995, KDD.

[6]  Philippe Fournier-Viger,et al.  Sequence Prediction using Partially-Ordered Episode Rules , 2021, 2021 International Conference on Data Mining Workshops (ICDMW).

[7]  Jerry Chun-Wei Lin,et al.  TKE: Mining Top-K Frequent Episodes , 2020, IEA/AIE.

[8]  Unil Yun,et al.  Efficient approach for incremental high utility pattern mining with indexed list structure , 2019, Future Gener. Comput. Syst..

[9]  Chia-Hui Chang,et al.  Efficient mining of frequent episodes from complex sequences , 2008, Inf. Syst..

[10]  Yun Sing Koh,et al.  A Survey of Sequential Pattern Mining , 2017 .

[11]  Jiadong Ren,et al.  Mining Frequent Intra-Sequence and Inter-Sequence Patterns Using Bitmap with a Maximal Span , 2017, 2017 14th Web Information Systems and Applications Conference (WISA).

[12]  Vincent S. Tseng,et al.  A novel methodology for stock investment using high utility episode mining and genetic algorithm , 2017, Appl. Soft Comput..

[13]  Ming-Yang Su,et al.  Applying episode mining and pruning to identify malicious online attacks , 2017, Comput. Electr. Eng..

[14]  Chenglie Du,et al.  Multi-Source Data Stream Online Frequent Episode Mining , 2020, IEEE Access.

[15]  Yun Sing Koh,et al.  Unsupervised Rare Pattern Mining , 2016, ACM Trans. Knowl. Discov. Data.

[16]  Antonio Gomariz,et al.  The SPMF Open-Source Data Mining Library Version 2 , 2016, ECML/PKDD.

[17]  Lina Fahed,et al.  DEER: Distant and Essential Episode Rules for early prediction , 2018, Expert Syst. Appl..

[18]  Fuzhen Zhuang,et al.  Online Frequent Episode Mining , 2015, 2015 IEEE 31st International Conference on Data Engineering.

[19]  Philippe Fournier-Viger,et al.  FMaxCloHUSM: An efficient algorithm for mining frequent closed and maximal high utility sequences , 2019, Eng. Appl. Artif. Intell..

[20]  Chaomin Huang,et al.  Mining High Average-Utility Itemsets Based on Particle Swarm Optimization , 2018 .

[21]  Farid Nouioua,et al.  Mining Partially-Ordered Episode Rules with the Head Support , 2021, DaWaK.