一种基于最大熵原理的社交网络用户关系分析模型

发布时间：2018-05-28 03:06

本文选题：社交网络 + 用户关系　；参考：《电子与信息学报》2017年04期

【摘要】：在社交网络的演化和发展过程中,用户之间关系的建立受到多种因素的共同作用。该文通过对社交网络中用户属性以及用户关系数据进行分析,旨在发现影响用户关系建立的关键因素。首先,针对用户关系建立的复杂驱动因素,分别从个人兴趣、好友关系、社团驱动3个方面提取影响用户关系建立的因素并定义相应的影响因子函数。其次,针对多种影响因素难以量化以及权值分配不确定等问题,以最大熵原理为基础构建用户关系分析模型,该模型在选择特征时具有不需要依赖于特征之间的关联性等特点,并能够量化各个因素对用户关系建立的驱动强度。从而挖掘影响链接建立的关键因素,分析用户关系发展态势。实验表明,该模型不仅能够量化各因素对链接建立的驱动强度,发现关键影响因素,而且可以对用户关系进行有效预测。
[Abstract]:During the evolution and development of social networks, the establishment of user relationships is influenced by many factors. Based on the analysis of user attributes and user relationship data in social networks, this paper aims to find out the key factors that affect the establishment of user relationships. Firstly, in view of the complex driving factors of user relationship establishment, the factors that affect the establishment of user relationship are extracted from three aspects of personal interest, friend relationship and community driving, respectively, and the corresponding factor functions are defined. Secondly, aiming at the difficulty of quantifying many factors and the uncertain distribution of weights, a user relationship analysis model is constructed based on the maximum entropy principle. The model has the characteristics of not dependent on the correlation between the features when selecting features. And can quantify each factor to establish the driving strength of the user relationship. The key factors that affect the link establishment are excavated, and the development trend of user relationship is analyzed. The experimental results show that the model can not only quantify the driving strength of each factor to link establishment and find the key factors, but also predict the user relationship effectively.
【作者单位】：重庆邮电大学网络与信息安全技术重庆市工程实验室;北京邮电大学北京市智能通信软件与多媒体重点实验室;
【基金】：国家973计划项目(2013CB329606) 国家自然科学基金(61272400) 重庆市青年人才项目(cstc2013kjrc-qnrc 40004) 教育部-中国移动研究基金(MCM20130351) 重庆市研究生研究与创新项目(CYS14146) 重庆市教委科学计划项目(KJ1500425) 重庆邮电大学文峰基金(WF201403)~~
【分类号】：TP393.09

【相似文献】