Adaptive Rule Update Method in Complex Event Process
Piyun Teng, Guanyu Li, Lei Su, Xinying Chen · 2017
Complex Event Process is a real-time data process technology based on the event process rule or event pattern, which was used to extract high level knowledge from a big amount of data, such as sensor data, stock data, etc. However, those rules, which are denied by the experts, are generally inaccuracy and static. And it is difficult to adapt to environmental changes. In this paper, to solve the problems, an adaptive rule updating method is proposed to adapt environment changing. Firstly, deleting error and unnecessary data in preprocess for improving the processing efficiency. And detecting abnormal rule, which is produced by monitoring the changes of Page Hinckley test(PHT) value. Secondly, updating the condition value of rule by adaptive cluster-based rule constraint method, and increasing a new better attribute by computing the effective value. However, rule constraint updating and new attribute increasing is depended on a given threshold representing the data size which is related to the current rule. If the variable value of data is bigger than the given threshold, it will compute the effective value of attribute, otherwise, the cluster-based adaptive constraint updating method is employing. This paper proposed an adaptive rule updating-based complex event processing framework for verifying the feasibility of the method.