中文摘要：

RFID数据的不可靠性是RFID事件处理的重要问题,传统的RFID数据清洗方法主要针对底层基本数据进行原始数据填补,但是有可能会造成大量无效数据冗余;针对RFID复合事件语义对原始数据的影响进行了研究,考虑了高层复合事件的语义信息对RFID原始数据所具有的限制约束作用,提出一种针对RFID复合事件漏检问题的处理方法：根据实际应用的语义信息形成带有时间约束的逻辑区域事件关系模型,通过关系模型找出符合时间约束的最有可能发生的复合事件以避免复合事件漏检.本文详细描述了该处理方法的处理策略并给出了算法描述.实验表明本文提出的方法有效减少了不可靠性处理时的数据冗余率;并在提高了复合事件合成准确率的同时降低了算法的处理时延.

英文摘要：

The unreliability of RFID raw data is an important issue of RFID event handling and traditional RFID data cleaning methods mainly focus on basic underlying data, by filling the missed reading data. However, these traditional methods may cause a lot of invalid data redundancy by filling a large amount of meaningless data. In this paper we identify the influence of RFID raw data on high-level composite RFID event synthesis. We consider the restrictions of the semantic information of high-level composite events over RFID raw data. And according to the characteristics of this relationship, we present a method focusing on missed reading problems by taking account of the semantic information of the high level application：the method first forms a logic regional event model according to the semantic of practical applications with time constraints. Then with the model we identify the most likely composite event to occur within the time constraints to avoid composite event missing. In this paper we describe the processing strategy and give the algorithm descriptions. Experiments showed that our method was effective in reducing the rate of the data redundancy when handling the unreliability problem, simultaneously improved the accuracy of the composite event synthesis algorithm and reduced the processing delay time.