Second-order local community common neighbor ratio and node correlation-based network connection edge prediction method

Abstract

The invention discloses a second-order local community common neighbor ratio and node correlation-based network connection edge prediction method. The method comprises the steps of building a network model, extracting first-order and second-order common neighbor nodes of two nodes without connection edges, and edges between the nodes to form a second-order local community, recording total numbers of the nodes and the edges of the community, and calculating an edge clustering coefficient, a simple harmonic average distance, connection edge density and a second-order local community coefficient; and calculating a Pearson product-moment correlation coefficient between the two nodes without the connection edges, calculating a common neighbor ratio, calculating a similarity index between the two nodes, calculating the similarity index between any two nodes without the connection edges in the whole network, arranging similarity scores between all node pairs without the connection edges according to a descending order, and taking the two nodes corresponding to first h indexes as predicted connection edges. According to the method, the common neighbor ratio between the nodes, the Pearson product-moment correlation coefficient and internal attributes of the local community are considered, and correlation information of the network is effectively utilized, so that the accuracy is relatively high and the prediction precision is relatively high.

Description

technical field [0001] The invention relates to the field of network and link prediction, in particular to a method for predicting network connection edges based on the proportion of common neighbors in a second-order local community and node correlation. Background technique [0002] With the vigorous development of Internet technology, human beings have entered a new era of Internet, and their understanding of the world has become more and more profound. Today, the Internet covers almost every aspect of our study, work, and life. Searching information online, finding jobs online, shopping online, etc. all involve the World Wide Web. More and more people use the Internet to obtain external information, so as to understand social dynamics. In real life, when we get along with people, we will form a network of interpersonal relationships, online shopping data will form a customized network of personal preferences, and transportation and travel will also form an intricate tr...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the disadvantages of incomplete acquisition of network information, low prediction accuracy, and poor prediction performance of existing methods for predicting network edges, and to obtain network information more comprehensively and improve the prediction performance of existing algorithms, the present invention proposes a A method of predicting network edges based on the proportion of common neighbors and node correlation in second-order local communities with high accuracy and high prediction accuracy

Images