符号网络的形成演化机制及其动力学研究

Signed network is a kind of network including edges with the property of positive or negative sign. Many real complex networks have opposite relationships, especially in social, biological and information fields. Using the sign properties of edges to analyse, understand and predict the topological structures, functions and dynamic behaviors of these real networks have important theoretical significance and practical applications, such as personalized recommendation, prediction of attitudes, public opinion control and so on. However, the relevant theories, as well as basic problems on signed network: the formation evolution mechanism and dnamical Processes on signed network have yet to be enriched.. Based on typical online signed social network, the project intends to investigate the formation and evolution on signed network, and the corresponding underlying social mechanism. The proposed research methods include：empirical analysis and theoretical modeling， covering different schools of qualitative and quantitative models from statistical physics to sociology. This project will focus on the critical phase transition phenomena between the emerging characteristics of macro pattern and the formation of the stable structure on the signed network, and explore the signed network adaptive dynamic behaviors.The proposed research will accomplish a series of modeling processes and computing techniques related to dynamic mechanism from micro individual behaviors to macro signed network structure evolution.. This project has important scientific significance. The research results will deepen our understanding on dynamical processes about complex signed system, enrich the basic theories on signed network and promote the potential applications.

符号网络是指边具有正或负符号属性的网络，广泛存在于信息、生物和社会领域。利用边的符号属性去分析、理解此类复杂网络的拓扑结构、功能、和动力学行为具有十分重要的理论意义, 并且对个性化推荐、态度预测、舆论管控等具有重要的应用价值。然而，符号网络的基础理论探索，符号网络的形成演化机制及其动力学过程等基本科学问题还有待充实。. 本项目拟通过实证分析与理论建模方法，从在线社会网络涌现的详细符号网络案例入手，结合从社会学到统计物理不同学派的定性、定量模型研究符号网络形成演化机制及其背后潜在的社会学机理。重点研究符号网络上宏观模式涌现特征和符号网络稳定结构形成的临界相变现象，探索自适应符号网络的动力学行为，实现从微观行为到宏观结构演化及动力学机制的建模过程及关键计算。. 研究成果将深化对复杂符号系统动力学过程的理解与认识，丰富符号网络的基础理论，推动符号网络的潜在应用。

符号网络是指边具有正或负符号属性的网络。本项目关注在线符号社会网络的形成演化机制，及其上的动力学过程。.主要研究内容和成果凝练如下：.1）揭示了符号社会网络演化背后的社会学机理是群体微观结构平衡和宏观统计scale-free规律联合驱动；.2）尝试用随机过程中几何布朗运动和分支过程建模符号网络的形成和演化动力学过程，进一步定量分析和预测符号网络上的动力学过程。.3）以仿真模拟及计算模型研究了多层耦合网络上的动力学过程，用超拉普拉斯矩阵谱分析、矩阵摄动理论的严谨数学分析证明多网络耦合推动了信息扩散和在线群体的共识。.4）提出了挖掘群体行为的时空模式关联算法。.5）提出了一种新的基于局部随机连边机制生成的网络模型、且能够描述真实世界在线社会网络的统计特征。. 研究结果深入探索了社交网络形成演化的机理， 揭示了多耦合社交媒体平台中的信息扩散机理、深化了对热点主题多路径社交媒体平台的传播机制的理解，而且为网络在线舆论调控提供了理论支撑。同时研究结论对在线社交媒体平台和公共信息发布系统的优化设计、提高用户体验、进行话题干预、舆情调控等具有实际应用价值。