基于众包(Crowd Sourcing)技术的定位机制研究.pdfVIP

基于众包(Crowd Sourcing)技术的定位机制研究 王文平1+ 1(南京大学 计算机科学与技术系,南京 210046) Research in positioning mechanism based crowdsourcing technology Wang Wenping 1+, 1(Department of Computer Science and Technology, NanjingUniversity, Nanjing 210046, China) + Corresponding author: Phn: +86-E-mail: 1226878649@ Abstract : The crowdsourcing model encourages users to add location information , raise own contribution , until the completion of dat a collection, forming a map database, provide services for the proposed location estimating users. Now the research about positioning mechanism using the crowdsourcing model is generally divided into two phases, at first, collection of the user through the cr owdsourcing model contribution data into the database, then acc ording to the data of the positioning request, matching estimation compared with database information. In this regard, I read two articles related references, first consider the method accords with the trad itional study, divided into three stages, plus a correction phase, the main consideration is device heterogeneity problems, and put forward the parameter fitting using EM algorithm, the unknown device linear transformation into a known training equipment; what second article con siders is using crowdsourcing to do trajectory estimation, it can reduce the influence of trace anomaly crowdsourcing user trajectory, distinguish the user is abnormal between normal users, and overcome the difficulty for trajectory space imbalances. Key words: Crowd Sourcing; mobile phone;positioning ;tag search;trajectory 摘 要:众包模式鼓励用户加入，通过众筹模式贡献自己的位置信息，等到数据收集完成后可以形成地图数据库，为提出定位需 求的用户提供位置估计服务。现在有关利用众包模式进行定位机制的研究一般分为两个阶段，先收集用户通过众包模式贡献的 数据，存入到数据库中，再根据数据的定位请求，对比数据库的信息进行匹配估计。对此，我阅读了两篇相关的参考文献，第 一篇考虑的方法符合传统研究，细化成三个阶段，加上了校正阶段，主要考虑设备的异质性带来的问题，并提出用EM 算法进 行参数拟合，将未知设备线性转化成已知的训练设备；第二篇文章考虑的是利用众包进行用户轨迹估计，其能够减少微量的异 常众包的用户轨迹的影响，区分用户为异常与普通用户，并克服众包因轨迹空间不平衡而带来的挑战。 关键词: 众包;移动手机;定位;轨迹 1 Device-clustering algorithm in crow