A Novel Complex-Events Analytical System Using Episode Pattern Mining Techniques

• Published in: Intelligence Science and Big Data Engineering. Big Data and Machine Learning Techniques Vol. 9243; pp. 487 - 498
• Main Authors: Tseng, Jerry C. C., Gu, Jia-Yuan, Wang, P. F., Chen, Ching-Yu, Tseng, Vincent S.
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 01.01.2015; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Complex event analytics; Data mining; Episode pattern mining; Graphics programming; Image processing; Multivariate sequence mining; Pattern recognition
• ISBN: 3319238612; 9783319238616
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-23862-3_48

Along with the rapid development of IoT (Internet of Things), there comes the ‘Big Data’ era with the fast growth of digital data and the requirements rise for gaining useful knowledge by analyzing the rich data of complex types. How to effectively and efficiently apply data mining techniques to analyze the big data plays a crucial role in real-world use cases. In this paper, we propose a novel complex-events analytical system based on episode pattern mining techniques. The proposed system consists of four major components, including data preprocessing, pattern mining, rules management and prediction modules. For the core mining process, we proposed a new algorithm named EM-CES (Episode Mining over Complex Event Sequences) based on the sliding window approach. We also make the proposed system integrable with other application platform for complex event analysis, such that users can easily and quickly make use of it to gain the valuable information from complex data. Finally, excellent experimental results on a real-life dataset for electric power consumption monitoring validate the efficiency and effectiveness of the proposed system.