A Community Discovery Approach for Noisy Networks

Abstract

The invention belongs to the technical field of social networks, and discloses a community discovery method suitable for noise networks, including: calculating the importance value of nodes in the network, establishing a set of core points and a set of boundary points; selecting core representative points to construct prior information; Select boundary representative points to construct prior information; combine prior information into the extremum optimization process; randomly divide the network into two parts with approximately equal number of nodes according to the topology to form the initial community structure; calculate the relationship between each node and the community module density Contribution value, move the node with the smallest contribution to another part for self-organization optimization, and repeat this self-organization optimization process until the module density value of the network no longer increases. The edges between the resulting two communities are removed until the module density value of the entire network reaches the maximum value. The invention effectively improves the accuracy of community division with relatively low cost, and improves the robustness of community division under noise environment.

Description

Technical field[0001]The invention belongs to the technical field of social networks, and particularly relates to a community discovery method suitable for noisy networks.Background technique[0002]Many networks in the real world, such as telephone networks, mail networks, and criminal networks, often contain incorrect or missing individual connections due to the difficulty in obtaining accurate and complete network structure information. Such networks are called noise networks. Most of the current methods of community discovery are to discover the community structure in the network based on the connection relationship between nodes in the network. Because these methods are completely dependent on the network topology and cannot be applied to noisy networks, when the proportion of noise in the network increases, the ability to discover the real community structure will decline rapidly; in the real network environment, partial prior knowledge of community division It can be learned. F...

AI Technical Summary

Problems solved by technology

[0003] To sum up, the problems existing in the existing technology are: the cost is high, and the obtained prior information may not have a strong guiding effect; the current community discovery methods cannot be applied to noisy networks and discover the real community structure. The ability of the network will decline rapidly with the increase of the noise ratio in the network; when obtaining prior information, high-quality prior information cannot be obtained with a small labor cost, which reduces the accuracy of community discovery

Images