一种用于移动计算的空间拓扑关联挖掘算法

摘要

在目前的移动智能系统中,由于使用现有挖掘算法存在大量的冗余候选项和重复计算量,制约着移动系统效率的提高；故提出一种基于事务支持项分离的拓扑关联规则挖掘算法,其适合在移动计算的多空间关系模式下,提取复杂的拓扑关联规则;该算法通过事务支持项分离法和上行搜索法,计算出空间拓扑关联规则,为移动用户提供决策支持；算法无需产生候选项和计算支持数:在仿真实验中,根据用户提供的不同支持度,与同类算法比较运行时间,实验结果表明其比现有算法快速而有效.%At present, since existing mining algorithms used in mobile intelligent system, which have some redundancy candidates and repeated calculated amount. And so these restrict the efficiency of mobile intelligent system. Hence, this paper proposes an algorithm of topology association rules mining based on transaction support items separation, which is suitable for extracting complex topology association in multi - spatial relation pattern of mobile computing. By transaction support items separation and up search, the algorithm computes spatial topology association rules to provide decision support for mobile customer, and it needn' t generate candidate and compute support. In the simulation experiment, according to different support given by user, we compare the runtime of the algorithm with similar algorithms, the result indicates that its efficiency is faster and more efficient than present similar algorithms.