计算机工程
計算機工程
계산궤공정
COMPUTER ENGINEERING
2014年
7期
67-74,92
,共9页
社交网络%社区检测%模块化%网络结构%跟踪%社交感知路由
社交網絡%社區檢測%模塊化%網絡結構%跟蹤%社交感知路由
사교망락%사구검측%모괴화%망락결구%근종%사교감지로유
social networks%community detection%modularity%network structure%tracing%social-aware routing
现有社区检测算法无法对社交活动和交互行为迅速发展的动态社交网络进行有效检测。为此，提出一种社区快速检测算法。使用现有网络知识确定的网络结构来更新网络社区，利用模块化技术自适应地检测和跟踪动态在线社交网络的社区结构。基于现实世界的动态社交网络对该算法进行测试，实验结果表明，使用该算法作为社区检测内核的社交感知路由策略，其性能要优于MIEN算法和Blondel算法。
現有社區檢測算法無法對社交活動和交互行為迅速髮展的動態社交網絡進行有效檢測。為此，提齣一種社區快速檢測算法。使用現有網絡知識確定的網絡結構來更新網絡社區，利用模塊化技術自適應地檢測和跟蹤動態在線社交網絡的社區結構。基于現實世界的動態社交網絡對該算法進行測試，實驗結果錶明，使用該算法作為社區檢測內覈的社交感知路由策略，其性能要優于MIEN算法和Blondel算法。
현유사구검측산법무법대사교활동화교호행위신속발전적동태사교망락진행유효검측。위차，제출일충사구쾌속검측산법。사용현유망락지식학정적망락결구래경신망락사구，이용모괴화기술자괄응지검측화근종동태재선사교망락적사구결구。기우현실세계적동태사교망락대해산법진행측시，실험결과표명，사용해산법작위사구검측내핵적사교감지로유책략，기성능요우우MIEN산법화Blondel산법。
The existing community detection algorithms cannot act on the dynamic social networks where social activities and interactions are evolving rapidly. To solve this problem, this paper presents a quick community-detection algorithm, which can quickly and efficiently update network communities by using the network structures identified from the previous network knowledge, and then an adaptive modularity-based method is proposed for identifying and tracing community structure of dynamic online social networks. To illustrate the effectiveness of the algorithm, it extensively tests the proposed algorithm on real-world dynamic social networks. The experimental results show that social-aware routing strategies employing the proposed algorithm as community detection core outperforms the MIEN algorithm and the Blondel algorithm.
