【摘要】 为改善传统同现模式挖掘算法基于内存计算的不足以及挖掘效率不高的问题,论文基于现有的同现模式挖掘算法,构建基于行人移动目标的粗细粒度结合的混合模型CFCMDCOP Graph,该混合模型有效保存了移动目标间的引发关系,以及对应实例间的时空关系,解决数据的存储问题。同时,论文给出相应的挖掘算法CFCMDCOP Graph Miner,可以提前对冗余模式进行剪枝。实验表明,该算法解决了数据存储问题,并提高了挖掘效率。更多还原

【Abstract】 In order to improve the deficiency of traditional co-occurrence pattern mining algorithm based on memory computing and the inefficiency of mining,this paper builds a hybrid model CFCMDCOP Graph based on the combination of coarse-grained and fine-grained pedestrian moving objects. The hybrid model effectively preserves the triggering relationship between moving objects,as well as the spatio-temporal relationship between corresponding instances,and solves the issues of data storage. At the same time,this paper presents the corresponding mining algorithm CFCMDCOP Graph Miner,which can prune redundant patterns in advance. Experiments show that the algorithm solves the problem of data storage and improves the efficiency of mining.更多还原