项目名称: 复杂网络模块结构与链接结构的统计建模及识别研究
项目编号: No.61202262
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: 赖大荣
作者单位: 东南大学
项目金额: 24万元
中文摘要: 本项目针对复杂网络模块结构分析和链接结构预测中的重要问题,采用随机统计建模的分析方法,分析模块结构分析和链接结构预测在统计建模上的内在联系,并对相应的建模框架和推理识别算法进行系统深入的研究。1)研究无向二元单部网络社团结构的建模分析框架及其推理算法,并基于分析框架设计有效确定网络中所含真实社团个数的模型选择准则;2)扩展无向二元单部网络的社团结构分析框架为适于分析有向网络、二部网络和加权网络一般模块结构的统计建模框架并设计相应的推理算法;3)研究链接预测的统计建模扩展分析框架及其推理算法;4)研究集成模块结构分析和链接结构预测的一致建模分析框架。本项目的研究成果,将有利于更好地理解复杂网络模块结构和链接结构模式的形成机制,对复杂网络拓扑结构模式的识别问题研究具有重要价值。
中文关键词: 复杂网络;模块结构;链接预测;聚类;
英文摘要: This project aims at important problems arising in modular structure analysis and link prediction of complex networks.Based on the statistical modeling relationship between modular structure analysis and link prediction, the project systematically and deeply investigates the modeling frameworks of modular structure analysis and link prediction and their corresponding inference algorithms, by employing methods of statistical modeling. Concretely, this project will: i) investigate the statistical modeling framework and inference algorithms for analyzing the community structure of unweighted binary unipartite networks, as well as the framework-based criterion for model selection to effectively determine the real number of communities in a network; ii) furter extend the framework and inference algorithms for community structure analysis of unweighted binary unipartite networks to the ones for analyzing more general modular structure of directed networks,bipartite networks and weighted networks;iii) investigate how to obtain extended statistical modeling framework for link prediction and its inference algorithms; iv) investigate the unified framework for simultaneously analyzing modular structure and link structure of networks.Results from this research will facilitate much better understanding of the mechanism for
英文关键词: Complex networks;Modular structure;Link prediction;Clustering;