Research Of Non-spontaneous Event Detect Based On Semantic Division

With the wide application of RFID technology, how to efficiently seek anddiscover associated events to complete deeper semantic composite event detectionbecome hotspot in RFID field. The Complex Event Processing as core of themiddleware technology attracts more and more attention from researchers. However,there exists a class of events in the composite event detection process, when theiroccurrence cannot be self-detect, until they are exact queried or expires. Such eventcalled passive event.In this paper, according to RFID non-spontaneous event detect research, wepropose a research of non-spontaneous event detect based on temporal semantic fornon-spontaneous event that is divided into forward reference and back reference.When processing non-spontaneous events, we improve the put forward the negationoperator to fusion with other types of operator for the query spanning tree.Based on the two semantic division senses, we preset shared counter forspontaneous queue and formulate relevant detection rules before the event detection.Respectively, we propose only-update the latest instance strategy and neutraldetection strategy in these two scenes. Only-update the latest instance can be used torelease expired non-spontaneous event instances timely and neutral detection caneffectively reduce the response delay.In the end, strategies proposed in this paper have been verified rationality andeffectiveness by experiment. Contrast shows that these strategies can significantlyreduce the time delay of the non-spontaneous event processing and memoryoverhead. It’s significant for processing large amount real-time RFID date flow.