基于Wi-Fi数据的校园社交网络分析

发布时间：2018-04-12 07:33

本文选题：Wi-Fi数据 + 语义轨迹　；参考：《北京邮电大学》2017年硕士论文

【摘要】：随着科技的迅猛发展,智能手机成为人们生活中娱乐和通信的必需品。而Wi-Fi功能已成为智能手机的标配,开启Wi-Fi时智能手机会自动地发送Wi-Fi数据包以探测周围环境中可用的Wi-Fi无线网络。这些Wi-Fi数据包中包含能够唯一标识用户的信息,获取、分析、挖掘这些Wi-Fi数据,对于研究智能手机用户的移动规律、揭示用户间的社交关系、检测用户的异常行为具有重要的意义。本文基于Wi-Fi探测器采集到的校园Wi-Fi数据集,研究学校内智能手机用户间可能存在的社交关系,进而挖掘与分析手机用户构成的社交网络。一方面,本文从语义轨迹的角度研究用户间的社交关系,在此基础上提出了两个语义轨迹相似度算法,分别为FA-STS(A Flexible Algorithm for Semantic Trajectory Similarity)算法和FP-STS(Frequent Pattern based Semantic Trajectory Similarity Algorithm)算法。相较于传统的基于GPS(Global Position System)的轨迹相似度算法,FA-STS算法灵活地控制了轨迹点匹配的维度,同时,通过进一步分析匹配轨迹点的特征,该算法可以判断用户间的具体关系;而FP-STS算法可以更准确地提取用户的语义轨迹模式,从而获得更高的准确率。经过实验验证,这两个算法能够更准确地衡量用户间的社交关系,提高算法的准确率。另一方面,本文提出RPC( Resident Population Classification)算法,用于在校园Wi-Fi数据集上提取不同类型的楼宇常驻人口的用户集合。在此基础上基于提出的语义轨迹相似度算法挖掘和分析了不同类型的楼宇常驻人口也即不同类型的人群组成的社交网络,并对其构成的不同的网络结构特征进行了分析。
[Abstract]:With the rapid development of science and technology, smart phone has become a necessity of entertainment and communication.The Wi-Fi function has become the standard for smartphones. When the Wi-Fi is turned on, the smartphone will automatically send Wi-Fi packets to detect the Wi-Fi wireless network available in the surrounding environment.These Wi-Fi packets contain information that uniquely identifies the user, acquires, analyzes and excavates the Wi-Fi data, which is useful for studying the mobile laws of smartphone users and revealing the social relationships between users.It is of great significance to detect the abnormal behavior of users.Based on the campus Wi-Fi data set collected by the Wi-Fi detector, this paper studies the possible social relations among the smartphone users in the school, and then excavates and analyzes the social network composed by the mobile phone users.On the one hand, this paper studies the social relations between users from the point of view of semantic locus, and then proposes two semantic locus similarity algorithms: FA-STS(A Flexible Algorithm for Semantic Trajectory similarity algorithm and FP-STS(Frequent Pattern based Semantic Trajectory Similarity algorithm.Compared with the traditional locus similarity algorithm based on GPS(Global Position system, FA-STS algorithm can control the dimension of locus point matching flexibly. At the same time, by further analyzing the characteristics of the matching locus points, the algorithm can judge the specific relationship between users.The FP-STS algorithm can extract the user's semantic locus pattern more accurately, so as to obtain higher accuracy.Experimental results show that the two algorithms can more accurately measure the social relationship between users and improve the accuracy of the algorithm.On the other hand, this paper proposes a Resident Population Classification (RPC) algorithm, which is used to extract the user sets of different types of resident population of buildings on campus Wi-Fi datasets.Based on the proposed semantic trajectory similarity algorithm, the social networks composed of different types of resident population (i.e., different types of population) are mined and analyzed, and the different network structure characteristics are analyzed.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92;TP393.18

【相似文献】