Social content risk identification method, device and facility

Abstract

The embodiment of the invention discloses a social content risk identification method, device and facility. The scheme comprises that feature dimension extension is performed on features of social content data to be identified through processing of a machine learning model of a stump structure, and then the obtained dimension-extended features are inputted to the deep machine learning model for prediction so as to identity whether there is a risk or not in the social content data.

Description

technical field [0001] This specification relates to the technical field of computer software, and in particular to a method, device and equipment for social content risk identification. Background technique [0002] With the rapid development of computer technology and Internet technology, more and more social activities are carried out based on the Internet, including but not limited to forum exchanges, instant messaging, online transactions and other activities. The corpus and other data are collectively referred to as social content here. [0003] In practical applications, some social content is risky and will affect normal users' online experience. In serious cases, it may threaten network security or even real social security, such as violence, terror, or politically-related inappropriate behavior or inappropriate speech. [0004] In the prior art, some keywords are often set based on experience, and risk identification is performed on social content through keyword ...

AI Technical Summary

Problems solved by technology

[0003] In practical applications, some social content is risky, which will affect the normal user's online experience, and may threaten network security or even real social security in serious cases, such as violence, terror, or politically-related inappropriate behavior or inappropriate speech, etc.

Images