User overlapping behavior-combined Microblog overlapped community partitioning method

Abstract

The invention discloses a user overlapping behavior-combined Microblog overlapped community partitioning method, relates to the field of Microblog user behavior assessment and overlapped community division, and in particular to an algorithm capable of displaying Microblog user overlapping behaviors and dividing Microblog overlapped communities. The method comprises the following steps of: firstly,constructing a relational network according to mutual concerning relationships of users, and assessing user overlapping behaviors such as user mutual-concerning behaviors, forwarding behaviors and mentioning behaviors; secondly, carrying out weight fusion according to the user overlapping behaviors so as to obtain relationship assessment functions between the users; and finally, weighting the relational network, fusing the user overlapping behaviors into a greed group expansion overlapped community division algorithm, and taking a Microblog weighting network as an initial network to improve community fitness functions. According to the method, user overlapping behaviors are embodied in community division, so that the overlapped community division quality in the Microblog network can be improved and the community structures embody that the users carry out communication intimately.

Description

technical field [0001] The invention belongs to the field of behavior evaluation of microblog users and division of overlapping communities, and specifically relates to an algorithm capable of displaying overlapping behaviors of microblog users and dividing overlapping communities of microblogs. Background technique [0002] Weibo, as one of the current important information dissemination platforms, has a large number of active Internet users, including public celebrities, government and enterprise officials, news media, and ordinary individual users. In the field of community discovery, community overlap is an important feature of real networks and has become a current research hotspot. Overlapping community discovery can find meaningful overlapping nodes, so as to reflect the network structure more realistically. Using complex network tools can solve real network problems, such as public opinion dissemination, power grid monitoring, network security, etc. Weibo social ne...

AI Technical Summary

Problems solved by technology

The method proposed in this patent can detect the overlapping community structure in the social network according to the keyword information of the nodes. Considering that the information is relatively single, the division results lack the reflection of the social network and user intimacy

[0005] The existing microblog network overlapping community division algorithm has the following shortcomings: (1) lack of consideration and reasonable quantification of comprehensive information such as user behavior; (2) the microblog network division results cannot reflect the typical characteristics of the microblog network

Images