面向移动时空轨迹数据的频繁闭合模式挖掘  

Frequent closed patterns mining for mobile trajectory data

作　　者：王亮[1] 汪梅[1] 郭鑫颖 秦学斌[1] 

机构地区：[1]西安科技大学电气与控制工程学院,陕西西安710054

出　　处：《西安科技大学学报》2016年第4期573-576,598,共5页Journal of Xi’an University of Science and Technology

基　　金：国家自然科学基金(61402360)

摘　　要：移动泛在感知设备的广泛普及为移动轨迹数据的大规模采集、存储与分析开拓了广阔的空间。通过对用户的移动轨迹数据进行分析挖掘,发现其中所蕴含的有价值的行为模式与特征,对于基于位置的服务(Location-based Service,LBS),城市交通管理,精准广告营销等领域均具有重要的价值。文中针对移动轨迹频繁模式规模过大、信息冗余问题定义了频繁闭合移动轨迹模式,以经典闭合序列模式挖掘算法为基础提出了适应于移动轨迹数据的频繁闭合模式Close Traj算法,分别通过对仿真数据与真实数据的实验测试,结果显示文中所提出的Close Traj算法对于频繁闭合移动轨迹模式挖掘问题具有较强的适用性,同时在运行效率方面具有显著优势。Thanks to the widespread popularity of mobile ubiquitous sensing devices,the acquisition, stor- age and analysis of large-scale mobile trajectory data have broad prospects for technology applications. By means of analysis and mining for users＇ mobile trajectory history ,we discover meaningful behavior pat- terns and characteristics behind the recorded trajectories. The above-mentioned discovered knowledge is of great value for location-based services, urban traffic management ,target advertising and many other areas. In this paper, aimed at the over-sized issue and information redundancy problem in frequent movement trajectory patterns, a conception of frequent close moving trajectory pattern is proposed. Moreover, based on classical closed sequential pattern mining algorithm, a frequent close pattern approach, namely CloseTraj algorithm, is devised under the condition of moving trajectory data. Based on the simulation and real data- set,the corresponding results show that our proposed CloseTraj algorithm has strong adaptability to the aforementioned problem with significant advantages in terms of operational efficiency.

关 键 词：移动轨迹 数据挖掘 频繁闭合模式 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]