欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对大规模网络的网络分析,本文提出基于社团为粒度的网络分割方法,以模块度作为评价准则,以节点网络属性作为启发式信息对网络进行分割,使得子图规模相当且具有社团群聚特征.社团子图规模相当使得经典的图论算法（如最短路径算法）充分发挥其作用;社团子图具有社团结构,使得子图之间连接边少,便于粒度分析.通过美国不同规模的城市交通网络的实验,证明了基于社团为粒度的网络分割的实用性,使得社团子网规模和社团子网数目都适合于经典最短路径算法.
英文摘要：
Due to large space demanding and time-consuming,the classic graph theory algorithms aren＇t suitable for solving some problems in the large-scale network.Most large-scale networks show community structure in the field of complex network.Sub-graphs of vertices have a higher density of edges within them while a lower density of edges between sub-graphs.Granular Computing imitates the human＇s thought of solving complex problems and granularity decomposition method is a good approach to simplify complex and large-scale problem.To solve large-scale network analysis,a community partition method of network based on granularity was proposed.By modularity for the evaluation criteria and network properties of node for heuristic information,it makes sub-graph scale similar and sub-graph with community structure.Due to sub-graph scale similar and small,the proposed method is beneficial effect on the classic graph theory algorithms（such as the shortest path algorithm） of network analysis.Owe to sub-graph with community structure,less connected edges among sub-graph is favorable for granular computing method.In U.S.city transportation networks of different scale,the result of experiments shows that the proposed method is effective.
同期刊论文项目
同项目期刊论文
位置：