项目名称: 在线社会网络中各主体动态行为间的相互作用分析
项目编号: No.61202179
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 计算机科学学科
项目作者: 李辉
作者单位: 西安电子科技大学
项目金额: 24万元
中文摘要: 随着Web2.0的快速发展,大型在线社会网络已经被广泛应用于网络推广、信息传播等方面。在线社会网络的发展使我们有机会面对海量的社交关系数据和信息,而对于这些信息的掌握可以使我们在市场推广、控制疫病、政策实施、通讯以及教育等诸方面受益匪浅。然而,我们对于这些社会网络的动态变化规律却仍然没有完整的认识,尤其是在个体行为、社团变化和网络整体规律的相互作用的理解上仍处于空白阶段。本课题中我们将利用偶图的理论,对社会网络中各层次主体(个体、社团和网络)动态行为的交互关系进行统一表示,并在此基础上将个体行为,社团变化和网络整体发展的研究统一成偶图中的边预测问题,从而挖掘这三者之间的联系和相互作用,并完善我们对于社会网络发展规律的认识。这种方法可以很好的解决个体、社团和网络动态行为的表示方法互斥的问题。此外,本课题还将系统的建立社会网络发展规律的多层次模型并进一步探索其应用于实际工作的途径。
中文关键词: 社会网络;数据挖掘;数据管理;知识发现;信息传播
英文摘要: With the rapid development and ubiquitous of Web 2.0 technology, online social networks are no long the simple network reflection of real social connections, they have been recently widely used in viral marketing, information propagaiton, etc. Social networks have become the largest media that information can spread through in an incredible speed. The development of these social networks makes it possible for us to acquire so much social connection data at unprecedented level. Mining useful knowledge from the data will benefit us much in many areas including marketing, virus spreading control, policy excution, education, etc. However, we have so little knowledge in the evolution pattern of social networks that we cannot make all the data fully utilized. Especially, the research towards the correlations and interactions between personal behavior, community evolution and network evolution is still in an early stage which requires much research efforts. To this end, we propose in this project a systematical approach which incorporates the mathematical description of personal behavior, community evolution and network change using a unique bipartite graph. Based on that, we turn the personal behavior learning, community evolution learning and network change learning into a unified problem, namely, link prediction.
英文关键词: social network;data mining;data management;knowledge discovery;information propagation