Internet link prediction method based on common neighbor node and community structure

Abstract

The present invention relates to an internet link prediction method based on a common neighbor node and a community structure. The method comprises the following steps: 1) initializing a network according to original network data, and acquiring a list Ln of node pairs without a link; 2) using a fast community detection algorithm BGLL to carry out community division on each node in the network, andassigning a unique community tag to each divided node; 3) according to the similarity index as shown in the specification based on the node community structure information and the RA index as shown in the specification improved according to the node common neighbor information, performing the summation to obtain the similarity Sx, y between any two nodes among the networks; and 4) according to the size of values of the similarity Sx, y, arranging node pairs in the list Ln from high to low, and selecting the first l node pairs in the list Ln, wherein the first l node pairs are a predicted linkthat is most likely to exist in the network or occur in the future. Compared with the prior art, the method provided by the present invention has the advantages that the link prediction accuracy is improved, the method can be applied to multiple networks, the prediction efficiency is improved, and the like.

Description

technical field [0001] The invention relates to the field of link prediction in network science, in particular to an Internet link prediction method based on common neighbor nodes and community structures. Background technique [0002] With the rapid development of network information technology represented by the Internet, human society has entered the era of complex networks. Human life and production activities increasingly depend on the safe, reliable and effective operation of various complex network systems. For example, the World Wide Web in the computer field, the power network in the energy field, the aviation network in the transportation field, the online dating network in the social field, and so on. As an interdisciplinary emerging field, network science has gradually formed and achieved rapid development. The link prediction problem in the network has benefited from the academic community's understanding of the importance of network science itself, and has al...

AI Technical Summary

Problems solved by technology

The first type of algorithm needs to obtain external information such as node attributes, but in many cases it is very difficult to obtain such information, and the practicability is poor

The second type of algorithm needs to construct the entire network ensemble, the amount of calculation is very large, and it will be difficult to deal with the network of thousands of nodes.

However, this type of similarity index based on common neighbor nodes uses very limited information and does not consider other topological features of the network such as community structure, so its prediction accuracy is limited

Images