随着网络和计算机科学的迅猛发展,处于网络结构洞位置的个人或团体从中可以获得丰富的信息利益和控制利益,结构洞占据者的挖掘得到了越来越多的关注.目前对结构洞占据者的挖掘研究主要集中于无权网络方面,但在实际网络中,权重对网络的性能也起着重要的作用.以加权网络为研究对象,将边权引入到中介中心性指标中,对现有的HIS算法进行改进,提出一种基于加权中介中心性的结构洞占据者挖掘算法(W_BCHIS).该方法首先利用社团发现算法求出加权网络的社团,接着将边权引入中介中心性指标,用加权后的中介中心性指标对节点的重要性进行初始化,通过对节点的重要性和节点占据结构洞的可能性不断地迭代直到收敛以挖掘出结构洞占据者.将W_BCHIS算法与其他算法在公用数据集和实际数据集上进行了实验对比与分析,验证了该算法的有效性.
With the development of various kinds of networks and computer science,the idea of what kind of position in the network could obtain the profits has been concerned by many people.The structural hole spanners may have access to abundant information and significant benefits.More and more attention has been paid on mining structural hole spanners.At present,researches on mining structural hole spanners are mainly for unweighted network.Actually,in a real network,the edge weight has a great influence on network performance.This paper proposes a novel weighted betweenness centrality algorithm to discover the structural hole spanners,named W_BCHIS,based on the existing algorithm HIS,for the weighted network.The algorithm W_BCHIS introduces the edge weight into the betweeness centrality function.This method first uses the discovery community algorithm to detect the weighted network community.Then the edge weight is introduced into the betweeness centrality function to initialize the impor-tance of nodes.Two functions to define the importance of nodes and the possibility of occupying the structural holes are calculated iteratly until they are converged.Finally,experimental results on the public datasets and real datasets show that the proposed algorithm W_BCHIS is better than the-state-of-art algorithms.