Complex Event Processing - Contextual Framework

There are two classes of data processing: batch and stream. During batch processing, a group of data elements is processed in entirety. A batch has crisp boundaries related to starting and stopping. Stream processing is a continuous process, where data elements are processed as they are received continuously. One type of stream processing is Complex Event Processing (CEP). CEP extends stream processing by supporting the combination of data elements from multiple streams and the construction of layers of abstraction. CEP has been extended to include semantic, probabilistic, predictive, and other forms of event processing. CEP systems utilize rules. Traditionally, rules have been written subject matter experts. Challenges related to manual rule generation include human error (for example including a value in Fahrenheit instead of Celsius), inability to handle emerging patterns, and scarcity of resources. An emerging area of research is the programmatic generation of rules.

Rules can be generated or refined, in multiple ways. Example approaches include Markov processes (Xuewei, Dongxia, Minhuan, & Xiaoxia, 2014), machine learning and statistics (Mehdiyev, Krumeich, Enke, Werth, & Loos, 2015), deductive reasoning (Margara, Cugola, & Tamburrelli, 2014), univariant time series analysis (Mousheimish, Taher, & Zeitouni, 2016), and clustering (Lee & Jung, 2016). The proposed area of research is the application of reinforcement learning to CEP rule generation.

References

Lee, O. J., & Jung, J. E. (2016). Sequence clustering-based automated rule generation for adaptive Complex Event Processing. Future Generation Computer Systems. doi:10.1016/j.future.2016.02.011

Margara, A., Cugola, G., & Tamburrelli, G. (2014). Learning from the past: automated rule generation for complex event processing. Paper presented at the Proceedings of the 8th ACM International Conference on Distributed Event-Based Systems, Mumbai, India.

Mehdiyev, N., Krumeich, J., Enke, D., Werth, D., & Loos, P. (2015). Determination of rule patterns in Complex Event Processing using machine learning techniques. Procedia Computer Science, 61, 395-401. doi:10.1016/j.procs.2015.09.168

Mousheimish, R., Taher, Y., & Zeitouni, K. (2016). Automatic learning of predictive rules for complex event processing: doctoral symposium. Paper presented at the Proceedings of the 10th ACM International Conference on Distributed and Event-based Systems, Irvine, California.

Xuewei, F., Dongxia, W., Minhuan, H., & Xiaoxia, S. (2014, 24-27 Aug. 2014). An Approach of Discovering Causal Knowledge for Alert Correlating Based on Data Mining. Paper presented at the Dependable, Autonomic and Secure Computing (DASC), 2014 IEEE 12th International Conference on.