基于复杂网络的Psor链路预测算法

A Psor Link Prediction Algorithm Based on Complex Network

复杂网络领域中,链路预测在网络演化规律的研究中被广泛应用,具有巨大的实际应用价值。针对现有的基于网络拓扑的链路预测方法存在预测精度偏低的问题,提出了一种基于复杂网络的Psor链路预测算法。该算法综合节点自身和邻居节点的度,定义了节点的Psor指数和Psor相似性指标进行链路预测。该算法能够全面考虑复杂网络的局部结构信息,更加准确地对复杂网络链路进行预测。仿真结果表明,Psor链路预测算法的预测精度相比8种经典的相似性算法的预测精度最高提升了37.96%。

In the field of complex networks,link prediction is widely used in the study of network evolution laws and it has great practical application value.For the problem of low prediction accuracy of existing link prediction methods based on network topology,a Psor link prediction algorithm based on complex networks is proposed.This algorithm integrates the degree of the node itself and neighbor nodes to define the Psor index of the node and Psor similarity index for link prediction.The algorithm can comprehensively consider the local structure information of complex networks and predict the links of complex networks more accurately.The simulation results show that the prediction accuracy of the Psor link prediction algorithm is up to 37.96%,higher than that of the eight classic similarity algorithms.