位置社交网络中移动对象社交关系发现方法研究

发布时间：2018-04-29 16:53

本文选题：时空交互 + 社交关系　；参考：《中国矿业大学》2017年硕士论文

【摘要】：社交平台与位置技术的紧密结合,促进了基于位置的社交网络的形成和发展。集成了GPS、无线网络、卫星定位等定位功能的智能设备为用户在社交平台上的位置标记、位置签到和位置共享提供了极大的便利。由此,产生了海量的、带有用户位置的数据。这些数据蕴含着大量的信息,能够为基于位置服务(如:用户社交关系发现、智能交通、旅游推荐、犯罪活动路径追踪等)的相关应用提供数据支持,迫切地需要研究人员对其进行全面的分析和有效的计算。本课题以基于位置的社交网络数据为研究对象,以用户社交关系强度计算和用户社交关系动态变化特征提取为研究目标,研究了灵活、全面的用户社交关系挖掘相关的理论和方法。本课题的主要研究工作如下:(1)本课题深入分析位置社交网络的时空交互性,充分考虑用户的行为特征和时空交互信息,提出了基于时空交互的用户社交关系强度计算方法。该方法通过对用户的时空交互情境进行建模,调整用户在不同情境下的交互权重,能够更加精确地计算用户之间的社交关系强度,更全面地发现用户之间的社交关系。(2)本课题深入分析位置社交网络中用户之间时空交互行为的变化特性,充分考虑用户行为特征和时空交互信息的变化情况,提出了基于用户行为特征漂移的动态社交关系强度计算方法。该方法通过构建用户行为特征漂移模型,利用时间切片技术,对用户之间的动态社交关系强度进行精确、全面和实时地量化计算,并提取用户社交关系的动态变化特征,有效地缓解了基于位置的社交网络数据的稀疏性问题。(3)本课题为了对基于位置的社交网络数据展开深入研究,扩充和丰富用户社交关系挖掘领域的相关理论和方法,加强用户社交关系挖掘理论和实践的结合,以基于时空交互的用户社交关系强度计算和基于用户行为特征漂移动态社交关系强度计算等研究成果为基础,设计并实现了时空社交关系计算原型系统。
[Abstract]:The close combination of social platform and location technology promotes the formation and development of location-based social network. Intelligent devices that integrate GPS, wireless network, satellite positioning and other location functions provide users with great convenience for location marking, location check-in and location sharing on social platforms. As a result, massive amounts of data with user locations are generated. The data contains a large amount of information that can support applications based on location-based services (such as user social relationship discovery, intelligent transportation, travel recommendations, criminal activity path tracking, etc.). There is an urgent need for comprehensive analysis and effective calculation by researchers. Based on the location-based social network data, the research object of this paper is to calculate the strength of user social relationship and extract the dynamic change feature of user social relationship. Comprehensive user social relationship mining related theories and methods. The main research work of this paper is as follows: (1) this paper deeply analyzes the spatio-temporal interaction of locational social networks, fully considers the behavior characteristics of users and space-time interactive information, and puts forward a method to calculate the intensity of user social relations based on spatio-temporal interaction. By modeling the user's space-time interactive situation and adjusting the user's interaction weight in different situations, the method can calculate the intensity of social relationship between users more accurately. Find out the social relationship between users in a more comprehensive way.) in this paper, we deeply analyze the changing characteristics of spatio-temporal interaction behavior among users in social networks, and fully consider the changes of user behavior characteristics and spatio-temporal interactive information. A dynamic social relationship strength calculation method based on user behavior drift is proposed. By constructing user behavior drift model and using time slice technology, the dynamic social relationship intensity between users is calculated accurately, comprehensively and in real time, and the dynamic change feature of user social relationship is extracted. In order to deeply study the location-based social network data, expand and enrich the relevant theories and methods in the field of user social relationship mining, which effectively alleviates the sparse problem of location-based social network data. To strengthen the combination of theory and practice of user social relationship mining, based on the research results of user social relationship strength calculation based on space-time interaction and dynamic social relationship strength calculation based on user behavior characteristic drift. A prototype system of spatiotemporal social relationship computing is designed and implemented.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.09

【参考文献】