Method and device for discovering important users on the basis of spreading networks

Abstract

The invention belongs to the technical field of social spreading analysis and specifically relates to a method and device for discovering important users on the basis of spreading networks. The method comprises a step 1 of constructing information spreading networks for activities required to be analyzed by means of clicking and sharing behaviors, and computing the spreading influence of each node in a single spreading network; a step 2 of regulating the spreading influence according to the number of primary interactive nodes of each node; a step 3 of computing the absolute influence of each node according to the effect of each node in multiple spreading networks and ordering the absolute influence; and a step 4 of normalizing data, transforming the absolute influence into relative influence, and ordering the relative influence to obtain the important users. The method prevents a possibility that a Pagerank algorithm plunges into an end node, increases computation speed, avoids a possibility of excessive iteration, comprehensively takes account of the effects of the node in multiple networks, and improves accuracy.

Description

technical field [0001] The invention belongs to the technical field of social communication analysis, and in particular relates to a method and device for discovering important users based on a communication network. Background technique [0002] With the continuous development of social networks, active information sharing and dissemination among people has become more and more common. Communicators publish the information they want to promote on the Internet through social media, and the information is seen by other individuals in their relationship network, which can cause multi-level secondary dissemination, and the exposure rate of information increases sharply. In the entire communication network, compared with other nodes, the key influence nodes have a great effect on the effect and scope of communication. Therefore, it is particularly important to locate the key influence nodes in the communication network in practical applications. [0003] When selecting key infl...

AI Technical Summary

Problems solved by technology

[0003] When selecting key influential nodes in the prior art, the algorithms often used include: 1. Degree centrality, that is, the more direct neighbor nodes of a node, the greater its influence. The disadvantage is that only the local information of the node in propagation is considered

2. The calculation of Pagerank is to regard the dissemination network as a directed graph. Each dissemination is recorded as a vote from the disseminator to the disseminated. Through iterative and recursive methods, the influence score of each node in the network is finally obtained. , the disadvantage is that in the propagation network, the connection between nodes is relatively sparse and scattered, the calculation efficiency of this algorithm is not high, and it is easy to fall into the possibility of terminal nodes

Images