Subjects: Computer Science >> Integration Theory of Computer Science submitted time 2018-04-24 Cooperative journals: 《计算机应用研究》
Abstract: For complex event detection in the mass real-time data from multiple sources, there is a problem of low accuracy and inefficiency in the triage of original event streams. This paper proposed a method of complex event detection based on an event tree. Firstly, defining dependencies between events, then take full account of the multi-dependencies between atomic events to generate atomic event trees and form the list with the dependent event trees, increasing the number of effective detection of complex event processing engines, such that the matching efficiency of event detection is improved. Meanwhile, this method reduces the memory consumption and improves the throughput of event detection. Simulation experiments and case studies demonstrate the advantages and feasibility of this method on massive data processing.