物联网用户的时空特征挖掘算法  

Spatiotemporal user profile mining algorithm in Internet of things

作　　者：于海宁[1] 方舟 马超 YU Haining;FANG Zhou;MA Chao(Research Center of Computer Network and Information Security Technology,Harbin Institute of Technology,Harbin 150001,China;Heilongjiang Province Cyberspace Research Center,Harbin 150001,China;School of Software&Microelectronics,Harbin University of Science and Technology,Harbin 150040,China)

机构地区：[1]哈尔滨工业大学计算机网络与信息安全研究中心,黑龙江哈尔滨150001 [2]黑龙江省网络空间研究中心,黑龙江哈尔滨150001 [3]哈尔滨理工大学软件与微电子学院,黑龙江哈尔滨150040

出　　处：《计算机集成制造系统》2020年第9期2429-2444,共16页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(61601146,61732022);国家重点研发计划资助项目(2016QY05X1000);黑龙江省自然科学基金资助项目(QC2018081)。

摘　　要：物联网中海量的人、物、信息往往具有时空特性,准确地获取用户时空特征是实现物联网个性化应用的前提条件之一。鉴于此,提出了物联网用户的时空特征挖掘算法,从用户全球定位系统轨迹中挖掘出频繁且呈现出异步周期性的时空运动模式。首先,利用全球定位系统点的时空特性将全球定位系统轨迹转化为热点区域序列。然后,采用一种基于模式增长的多最小支持度的异步周期的序列模式挖掘算法,按照多最小支持度,深递归地挖掘出异步周期的序列模式。最后,通过实验证明了所提算法的有效性和准确性。Internet of Things(IoT)encompasses a huge number of users and physical entities,which usually have temporal and spatial features.User spatiotemporal profile is the key prerequisites to build personalized applications in IoT.Therefor,a user spatiotemporal profile mining algorithm was proposed to discover travel patterns from user Global Positioning System(GPS)trajectories.A GPS trajectory was transformed into a sequence of Regions of Interest(ROI)based on spatial and temporal property of GPS points.Then a pattern-growth mining algorithm was proposed to mine asynchronous periodic sequential patterns with multiple minimum item supports,which were not only occurring frequently,but also appearing periodically.The experimental results showed the efficiency and accuracy of the proposed algorithm.

关 键 词：物联网 用户特征 时空特性 GPS轨迹 序列挖掘 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]