Multi-layer iterative screening method and device for relation network

Abstract

The invention discloses a multi-layer interactive screening method and device for a relation network. The method includes the steps that a social relation network diagram is obtained; according to theattribute value of each node, multiple seed nodes of the social relation network diagram are screened; according to the average value of the attribute values and side attribute values of all the nodes, initial screening values are obtained; according to the initial screening values, the seed nodes are expanded for multiple times, and expanding is stopped when expanding stopping conditions are metso as to obtain a sub-diagram of the social relation network diagram; the number of nodes of the sub-diagram, whether the number of the nodes of the sub-diagram is larger than a target expanding number is judged; if yes, the sub-diagram is output, otherwise the expanded nodes are used as seed nodes, the initial screening values are regulated to continue screening until the number of the nodes ofthe sub-diagram is larger than the target expanding number. The method effectively improves the accuracy and the covering rate of a recommended algorithm, and enables the sub-diagram to be more visualand have higher reference value.

Description

technical field [0001] The invention relates to the technical field of computer networks, in particular to a multi-layer iterative screening method and device for relational networks. Background technique [0002] At present, with the continuous development of Internet technology and social networking sites, data mining has received great attention, and the problem of information overload has become more and more prominent. An important problem faced is that accurate screening of relational networks can help users to better conduct network analysis and information search. [0003] In the related art, in order to realize the screening and compression of the social relationship network, the following methods are generally adopted in the industry: greedy algorithm, a two-terminal network simplification method, etc. Although these methods have many advantages, such as considering the attribute value of the node, the relationship between the nodes, etc. However, the core influen...

AI Technical Summary

Problems solved by technology

[0003] In related technologies, in order to realize the screening and compression of social network, the industry generally adopts the following methods: greedy algorithm, a double-ended network simplification method, etc. Although these methods have many advantages, such as considering the attribute value of the node, the However, it does not consider the core influence of nodes, the closeness of cooperation between nodes, the overall structure, and the overall balance between structural nodes. For large-scale social networks, these algorithms not suitable

Images