Identity and motivation identification method and system for atypical media account

Abstract

The invention discloses an identity and motivation identification method and system for an atypical media account, relates to the field of public opinion big data, and the method includes obtaining data published by each atypical media account of each Internet platform, generating a basic database, and account according to the basic database, respectively calculating the number of original sentences and participation times of each atypical media account for the same event; distributing each atypical media account in each quadrant under the two-dimensional coordinate system according to the number of original sentences and participation times; and according to the quadrant where each atypical media account is located, identifying the identity and motivation of each atypical media account, so that the identification efficiency, accuracy and practicability are improved, and the identification difficulty is reduced.

Description

technical field [0001] The present invention relates to the field of public opinion big data, in particular to a method and system for identifying the identity and motivation of atypical media accounts. Background technique [0002] With the rapid development of mobile Internet technology and the continuous improvement of information infrastructure, the basic structure of the Internet and the application form of the Internet have been further upgraded. During this process, new changes have taken place in the concepts and habits of using the Internet. During this period, a significant change was that more and more Internet users (individuals or organizations) voiced their voices through Internet platforms such as Weibo, WeChat, Toutiao, Zhihu, Douyin, and Kuaishou, becoming a type of SARS besides institutional media. Sexual media. [0003] The connotation of this type of media refers to individuals or organizations that use network technology, equipment and applications to ...

AI Technical Summary

Problems solved by technology

[0015] (1) There is more theory than practice, most academics conduct theoretical experiments through offline data of individual cases, the reality is not generalizable enough, and the application is relatively narrow;

[0016] (2) The timeliness is poor, and the identification cannot be completed in time;

[0017] (3) The operation is difficult, requiring researchers to have in-depth application capabilities in natural semantic analysis and machine learning;

[0018] (4) The accuracy is not high. The premise of the above identification method is that the data of different Internet media platforms have similar formats and are stored in the format of key-value pairs, that is, each item corresponds to a value, which provides the possibility for similarity calculation As long as the items with the same basic information of users on two platforms are specified, the similarity can be calculated, ignoring the natural similarity of basic information of users on different platforms