Adaptive Rule Model Based Complex Event Processing

IoT applications are based on wireless sensing devices which achieving things-be-interconnected by wireless sensing devices.This results in real-time,larg amounts of the data,heterogeneity of the data and so on.For this reason,the complex event processing technology is proposed.This paper analyzes the problems of currently complex event processing:complex event processing is limited to effective rule matching from event process rules,but which are whether adapt to environmental changes or not is seldom involved yet.Secondly,existing complex event processing is seldom considered in rule matching.The study works of this paper are focused on above-mentioned two problems in com-plex event processing system.(1)Aiming at the problem of maladjustment to environment changes in rules defini-tion,an adaptive event process rule model is proposed based on anomaly detection and classification rule learning.Via the results from rule evaluation,rules dissatisfied current environmental changes should carry out a self-relearning process from happened events.(2)Aiming at the inability of rules matching to uncertain data and event complexity,a method of uncertain model integrated process of rule matching is proposed,which can fil-ter matching results from rule matching by event type uncertainty model,so that the result accuracy of complex event processing is promoted and the storage pressure of intermedi-ate results is reduced.By storing matching nodes in multiple conditions,all unnecessary matching are reduced.Finally,the experimental results reveal that the proposed concrete method corre-sponding to two problems are effective and efficient.