Overlapping community discovery method and device based on streaming analysis

Abstract

The invention belongs to the technical field of network overlapping community discovery, and discloses an overlapping community discovery method based on streaming analysis, and the method comprises the steps: calculating the mode of node degree in a network as a threshold value, and then carrying out the randomization of the sequence of edges in the network through employing a Fisher Yates shuffling algorithm; mining the community structure hidden in the network by taking information such as degree information of nodes, contribution degree of the nodes to communities and number change of connecting edges among the communities as standards of node division. The invention further discloses an overlapping community discovery device based on streaming analysis. According to the method, the precision of carrying out community discovery tasks in a large-scale network is improved, the algorithm is expanded to overlapping communities, and the communities can be rapidly discovered in the large-scale network under linear time and space complexity.

Description

Technical field[0001]The invention belongs to the technical field of network overlapping community discovery, and in particular relates to a method and device for overlapping community discovery based on streaming analysis, which is used in large-scale networks.Background technique[0002]With the rise of social networks and the Internet of Things, the scale of networks has become larger and larger, and many existing networks have grown to have tens of millions of nodes and billions of edges. It is very important to quickly find the hidden community structure in these large-scale networks, and many traditional methods cannot handle such large-scale networks in terms of computing time and memory consumption. The literature (Hollocou A, Maudet J, Bonald T, et al. A Streaming Algorithm for GraphClustering[C].neural information processing systems, 2017:1-10.) proposes a linear streaming large-scale network community discovery algorithm SCoDA , The algorithm has linear time and space compl...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the problem that the accuracy of SCoDA algorithm community discovery is not high, and nodes can only be divided into single communities, the present invention proposes an overlapping community discovery method and device based on flow analysis, under the condition of maintaining linear time and space complexity, Improves the quality of community partitioning in large-scale networks and extends the algorithm to domains of overlapping communities

Images