Social network immune algorithm based on network high-order structure

Abstract

The invention belongs to the technical field of computer science, and discloses a social network immune algorithm based on a network high-order structure. The algorithm comprises the following steps: excavating a motif in an original network; reconstructing a topological structure of the network, and manufacturing a plurality of random networks; obtaining a Z-score value of each die body through a random network; selecting the motif with the maximum Z-score value as a target motif, and generating an adjacent matrix based on the target motif; setting the weight of the connecting edge according to the value of the connecting edge in the adjacent matrix based on the target die body, and re-weighting the original network to obtain a weighted network; obtaining the intermediary centrality of the connection edges based on the weighted network; sorting the values of the intermediate centralities of the connecting edges from large to small, and selecting the first k connecting edges to deploy immune resources, so as to realize immune operation. According to the algorithm, the problem that the network immune algorithm neglects the high-order connectivity of the network and damages the integrity and functionality of the network in the prior art is solved.

Description

technical field [0001] The invention relates to the technical field of computer science, in particular to a social network immune algorithm based on a network high-order structure. Background technique [0002] As we all know, the development of science and technology, especially information technology, has made human society more and more closely connected, but it also brings huge risks. In order to deal with the large-scale dissemination of negative information in the network, the most recognized method is network immunity, that is, by "immunizing" some social network nodes and other structures, so as to prevent other nodes in the network from being affected by bad information. The best network immunization algorithm can minimize the number of affected nodes when immunizing the same proportion of nodes, edges or other network structures. [0003] In the research of network immunity, the connection degree of nodes, the influence of nodes and various centralities are often ...

AI Technical Summary

Problems solved by technology

First, there are some hidden and important structures in the social network that will affect the importance of nodes or edges, while previous immune algorithms often ignore this point; second, node betweenness and node connectivity are based on network Metrics of nodes In real social network immunization, immunization of important nodes often seriously affects the normal activities of adjacent and even all other nodes and will destroy the integrity and functionality of the network

Images