复杂社会网络节点重要性可拓聚类动态分析方法

目前大多数复杂网络关键节点影响力识别方法都是建立在静态分析基础之上的,缺乏对节点动态变化的考量。采用可拓聚类方法来量化分析社会网络的动态变化,以多属性决策法测量各节点静态重要性属性值;分别划分等级,取经典域和节域,计算各节点在对应属性上的关联度,并用熵权法取得综合关联度;分析其在同一社会网络不同时间点的变化,给出变动的量和方向。该方法不仅能刻画出节点重要性的变化趋势,而且给出了变化程度。最后,通过实例验证了该方法的可行性和有效性。

EXTENSION CLUSTERING DYNAMIC ANALYSIS METHOD FOR IMPORTANCE OF COMPLEX SOCIAL NETWORK NODES

At present,the identification method of the influence of key nodes in complex networks is all based on static,which is lacking in consideration for dynamic analysis. In this paper,the expansion clustering method was used to quantitatively analyze the social network under the dynamic change. The value of each node importance was measured by multi-attribute decision. Then,the grade was divided respectively,the classical domain and acceptable domain were taken,and the correlation degree of each node on the corresponding attribute was calculated. Meanwhile,the entropy weight was used to obtain comprehensive correlation degree. We analyzed the node importance of the same social network at different time points,and gave the amount and direction of change. This method can not only describe the trend of changing,but also work out the amount of it. At last,an example is given to verify the feasibility and effectiveness of the proposed method.