基于资源传输节点拓扑紧密性的链路预测方法

针对现有基于网络拓扑结构的局部相似性RA指标未考虑传输节点拓扑紧密性的问题,提出一种节点拓扑紧密性指标及链路预测方法。根据多跳节点资源传输情况确定重要传输节点,基于传输节点周围拓扑集聚程度对拓扑紧密性进行量化,并根据传输节点紧密性对共同邻居传输资源量的影响刻画节点间相似性。实验结果表明,该方法具有较高的普适性,所提相似性指标适合于Precision标准,与CN、AA和CAR等现有相似性指标相比,具有较高的预测精度。

Link Prediction Method Based on Topological Tightness of Resource Transmission Nodes

To address the problem that the existing local similarity RA index based on network topology does not consider the topological closeness of transmission nodes,this paper proposes a node topology closeness index and link prediction method.The important transmission nodes are determined according to the resource transmission details of multi-hop nodes,and then the topological closeness is quantified based on the analysis of the topological clustering degree around the transmission nodes.Finally,the similarity between nodes is described by using the impact of the closeness of the transmission nodes on the transmission resources of the common neighbors.Experimental results show that this method has high universality,and the proposed similarity index is suitable for the Precision standard,and has higher prediction accuracy compared with the existing similarity indexes such as CN,AA and CAR.