摘要
为从海量机动车行驶数据中找出车辆的伴随车辆信息,本文依据伴随车辆的行为特点及关联挖掘原理提出了一种基于频繁项集的伴随车辆检测算法。通过分析伴随车辆的行驶特点,设定频繁项集的支持度和时间阈值,实现了从单项集到多项集的迭代过程。利用流数据处理方法,添加了针对伴随车辆检测的流数据处理方法,最终使算法可以快速的在海量数据中检测到车辆的伴随车辆信息。经实验验证,本算法可以快速正确的检测出车辆的伴随车辆信息。
To detect car’s accompany vehicle in mass vehicle driving data, this paper analyzes the features of ac-company vehicle and principles of association rules mining. Finally, an algorithm based on frequent itemset is proposed in this paper. Support and time threshold of frequent itemset is set up to realize the process from single-itemset to multi-itemset with analysis car’s driving features on the road. The stream data is considered and a method is come up to solve the problem of new data in detecting accompany vehicle. Eventually, the result of the given car’s accompany ve-hicle can be detected quickly in mass data. Through the experiments, the proposed algorithm can get the accompany vehicle’s information quickly and right.
出处
《软件》
2016年第4期69-73,共5页
Software