Link prediction method based on bayes estimation and common neighbour node degree

Abstract

The invention discloses a link prediction method based on bayes estimation and common neighbor node degree. The method comprises the steps of establishing a network model; taking any two nodes which are not connected directly as seed nodes and calculating a probability that a connected side exists or do not exists between the nodes; calculating the probability that the connected side is generated or is not generated between the two nodes according to degree information of an intermediate node of a path with a length of 2 or 3 between the two nodes; calculating a likelihood value of each intermediate node of the path with the length of 2 or 3 between the two nodes according to the bayes estimation and the common neighbor node degree, wherein a similarity score is the sum of the likelihood values of all intermediate nodes; and traversing a network, obtaining the similarity score between any two seed nodes through adoption of the method, sorting all seed node pairs according to the similarity scores in a descending order, and taking the node pairs corresponding to the former B score values as predicted connected sides. According to the method, on the basis of the bayes estimation and through combination of the common neighbor node degree, different intermediate nodes in the partial path between the two nodes have different importance and a prediction effect of the algorithm is good.

Description

technical field [0001] The invention relates to the fields of network science and link prediction, in particular to a link prediction method based on Bayesian estimation and common neighbor node degree. Background technique [0002] Complex systems in real life can be studied using complex networks. Nodes in the network represent individuals in the complex system, and edges represent the interrelationships between nodes in the system. Link prediction is one of the important research fields of complex networks, because link prediction can predict the links that may be generated between nodes during the evolution of the network, so the evolution trend of the network can be predicted in advance, and it can be judged The "ghost edge" that does not exist in the network can better help researchers study the internal laws of the network. [0003] The link prediction problem has received extensive attention from researchers. In comparison, the link prediction algorithm based on ne...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the problem that the existing link prediction method based on the degree of common neighbor nodes only considers the intermediate nodes of the path whose path length is equal to 2, and only considers the degree of these nodes, which results in low prediction accuracy, the present invention proposes a A Link Prediction Method Based on Bayesian Estimation and Common Neighbor Node Degree with High Accuracy

Images