基于Spin-glass模型的多角度识别可控的社区发现算法-计算机应用研究.DOCVIP

基于改进PageRank算法和Spin-glass模型的多角度识别可控的社区发现算法 石 云1,2,3 , 陈 钟2,3, 孙 兵2,3,4 (（1. 六盘水师范学院 计算机科学与信息技术系，贵州 六盘水，553004； 2. 北京大学 信息科学技术学院，北京 海淀区 100871； 3. 北京大学网络与软件安全保障教育部重点实验室，北京 海淀区 100871； 4. 广东海洋大学 信息学院，广东 湛江，524088)） 摘要:为了针对如何有效解决社区发现算法中的社区重叠问题，以及如何改善社区扩张模式，提出了一种基于改进PageRank算法和Spin-glass模型的多角度识别可控的社区发现算法(PRSGMFCA)被提出，。首先通过优化的算法对每一个节点排序,，确定其中的核心节点；；再通过基于模型的多角度识别模块度优化局部社区的扩张模式，解决传统模块度在分辨率极限方面的束缚和影响；同时同时在发现过程中，引入了改进的贪心迭代算法发现局部社区；最后，最终实现准确地发现重叠结构与节点。经过在计算机模拟网络与真实网络环境下应用比较分析， 发现和传统的技术方案相比具有较好的稳定性与正确率，,并且其算法复杂度也在可以接受的范围内。 关键词:社区发现；改进PageRank算法；Spin-glass模型；多角度识别可控；贪心迭代算法 中图分类号: TP309.7 文献标志码: A A Discovery algorithm Of of multi-angle identify controllable based On on improved PageRank Algorithm algorithm Aand Spin-Glass Models models SHI Yun1,2,3，CHEN Zhong2,3，Sun Bing2,3,4 (1. Liupanshui Normal College University, Department of Computer Science and Information Technology, Guizhou, Liupanshui, 553004, China; 2. School of Electronics Engineering and Computer Science, Peking University, Beijing, Haidianqu, 100871, China; 3. Network and software security for Ministry of Education Key Laboratory, Peking University, Beijing, Haidianqu, 100871, China; 4. College of Information, Guangdong Ocean University, Guangdong, Zhanjiang, 524088, China) Abstract: In order to effectively address solve the problem of communities overlap in community discovery algorithm, and improve the community expansion model, the paper propose a discovery algorithm of multi-angle identify controllable based on improved PageRank algorithm and Spin-glass models is proposed: . Firstly, utilize optimized algorithm to sort each node, determine which of the core node; Secondly, utilize multi-angle identification module degree based on the model to optimize the expansion model of local community, and solve the shackles and influence of traditional modularity in terms of the resolution limit, meanwhile in the discovery process, introduce an improved greedy iterative algorithm to discover local communities