A Context Sensitive Method for Complex Event Processing

With the rapid development of Cyber-Physical Systems (CPS), complex event processing, a crucial technology in the real-time processing of big data streams, is garnering increasing attention. This paper takes complex event processing as the foundation, leveraging fuzzy ontology to model the context of historical events. It accomplishes data partitioning via context clustering and learns corresponding Bayesian network models tailored to different datasets. During real-time prediction, an appropriate Bayesian network model can be chosen for predictive analysis in accordance with the context of the current event. Experimental results demonstrate that this approach can efficiently handle event data streams and exhibits excellent prediction performance.