Method and device for detecting junk posts based on occurrence rule of posts

Abstract

The invention provides a method and device for detecting junk posts based on an occurrence rule of posts in a community network. The method comprises the following step: a, recognizing the posts, and judging whether the posts are junk posts according to the content characteristics and the occurrence rule in one or a plurality of community networks. Preferably, the step a comprises the following steps: a1, recognizing the posts according to the preset semantic rule, and extracting the content characteristics of the posts; a2, inquiring the occurrence rule of the posts in the community network according to the content characteristics of the posts; and a3, judging whether the posts are junk posts based on a first preset rule according to the occurrence rule of the posts in the community network. In the prior art, a large number of repeated posts occurring in the community network can not be detected, which is caused by carrying out the dirty word matching or semantic analysis on a single post individually in general, and compared with the prior art, the method and the device can be used to improve the judgment accuracy of the junk posts.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to a method and device for detecting spam posts in a community network. Background technique [0002] With the continuous development of Internet technologies, social network services (SNS, Social Network Service) are becoming more and more popular and gradually become a part of people's daily life. However, the proliferation of spam posts based on social networks and the resulting interference to real useful information have always been the disadvantages of the vigorous development of social networks. For this reason, in order to effectively suppress the generation of spam information in social networks, the prior art at least includes the following methods for filtering spam content in posts in social networks: [0003] (1) Dirty word matching, that is, before the user publishes the post on the social network, after at least one dirty word filtering, the words in the content...

AI Technical Summary

Problems solved by technology

[0006] It can be seen that the prior art is based on the judgment of the content of a single post to realize the shielding of the spam content in the post, that is, the prior art is only limited to filtering the content of the post within the scope of a single post, so it cannot be applied to Such a situation: the spam characteristics of a single post are not obvious or hidden (such as soft text posts), but in fact there are a large number of duplicate posts that need to be deleted in the entire social network

Images