微内容传播的群体行为感知与角色挖掘关键技术研究

项目名称： 微内容传播的群体行为感知与角色挖掘关键技术研究

项目编号： No.71461013

项目类型： 地区科学基金项目

立项/批准年度： 2015

项目学科： 管理科学

项目作者： 柳军

作者单位： 江西农业大学

项目金额： 34万元

中文摘要： 微内容的泛在化、社区化传播是网络舆情形成的重要原因，复杂网络和信息挖掘相关研究成果可供参考。用户的群体行为对网络结构和舆情演变的影响极为显著，具有重要研究价值。前期工作微内容网络传播与免疫策略研究,发现微内容具有多重媒体属性的信息特征，经过传播节点各种微行为的互动会导致观点突变和新行为的涌现。受节点角色的微观动力学支配，网络的群体行为使得传播网络社区化，往往会出现观点多元极化现象。拟在前期阶段性成果的基础上，进一步分析社区化网络传播者和群体传播行为的特点和规律，在理论、算法和应用三个层面开展研究，基于意见动力学建立感知群体行为的优化模型，引入合适的参变量，寻找群体行为形成的临界点；采用统计物理的理论和方法研究节点角色挖掘的算法，探讨角色驱动下的群体行为演化和控制模型，并在真实大规模网络中测试和评估所提模型与算法的表现，验证其应用价值，以期提出自媒体环境下网络舆情管理的新机制和新策略。

中文关键词： 网络舆情；复杂系统建模；数据挖掘；危机信息管理；知识发现

英文摘要： The ubiquitous and community communication of Micro-content is the important reason for the formation of network public-opinion in, Some research result can be referenced.The group behavior of users is more Significant than individual nodes to the network structure and opinion evolution.So it has have more value to related research.?We found that micro-content has the information characteristics of multiple media properties through our previous research.It will lead to the point mutation and the emergence of new Behavior by the nodes' interaction of micro-behaviors. Dominated by micro-dynamics of node role, the group behavior makes the transmission network community, and the opinion often appears multi polarization phenomenon. On the basis of previous works, we will further analysis the characteristic and law of communicators and group communication behavior on social networks. We prepare to carry out the research in the theory, algorithm and application three aspects. We wish establish the optimized model of recognition group behavior based on opinion dynamics, to find the critical point of group behavior formed by introducing appropriate variables, and research on node role mining algorithm by using the theory and methods of statistical physics. We will explore the evolution and control model of group behavior under role driving, test and evaluate the performance of the model and the algorithm in a real large-scale network.and then verify the application value. At last, we will put forward some new mechanism and strategies of network public opinion management in the time of we-media.

英文关键词： Network Opinion;Complex System Modeling;Data Mining;Crisis Information Management;Knowledge Discovery