Mining Interventions from Parallel Event Sequences
Discovering temporal patterns from sequence data has been an important task of data mining in recent years. In this paper a novel temporal pattern, Intervention, is proposed to capture the partial ordering relations in parallel event sequences. It is demonstrated that Intervention is essentially a deviation of generalized Markov property holding in parallel event sequences. A measure to evaluate the degree of such deviation, Intervention Intensity, is suggested, which has an important mathematical property, non-symmetry. As a result, an algorithm called MIPES for mining interventions is proposed. The time complexity of MIPES is of O(m 2) and is independent of data size, where m is the number of event types and is far smaller than the data size in practice. The experimental results show MIPES is applicable and scalable.
KeywordsParallel Event Sequence Intervention
Unable to display preview. Download preview PDF.
- 1.Unnikrishnan, K.P., Uthurusamy, R. (eds): Temproal Data Mining Workshop Notes. In: 8th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2002) (2001)Google Scholar
- 2.Vautier, A., Cordier, M., Quiniou, R.: An Inductive Database for Mining Temporal Patterns in Event Sequences. In: International Joint Conference on Artificial Intellegence (2005)Google Scholar
- 5.Antunes, O.: Temporal data mining: An overview. In: 7th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 2001) (2001)Google Scholar
- 8.Laxman, S., Sastry, P.S., Unnikrishnan, K.P.: Discovering Frequent Episodes and Learning Hidden Markov Models: A Formal Connection. IEEE Transactions on Knowledge and Data Engineering 19(11), 742–758 (2005)Google Scholar
- 10.Agrawal, R., Srikant, R.: Fast Algoritms for Mining Association Rules. In: VLDB, pp. 487–499 (1994)Google Scholar
- 11.Agrawal, R., Srikant, R.: Mining sequential patterns. In: ICDE, pp. 3–14 (1995)Google Scholar
- 14.Keogh, E., Pazzani, M.: An Enhanced Representation of Time Series Which Allows Fast and Accurate Classification, Clustering and Relevance Feedback. In: Knowledge Discovery in Databases and Data Mining, pp. 239–241 (1998)Google Scholar
- 15.Lesh, N., Zaki, M., Ogihara, M.: Mining Features for Sequence Classification. In: The 5th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 1999) (1999)Google Scholar
- 16.Ketterlin, A.: Clustering Sequences of Complex Objects. In: The 3th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD 1997) (1997)Google Scholar