A detection method of microblog social robot based on deep neural network

Abstract

The invention discloses a microblog social robot detection method based on a deep neural network. The method includes constructing a manually marked data set, extracting classification features of user data, and constructing a residual neural network, a bidirectional gated recurrent unit and an attention mechanism. A deep neural network model for Sina Weibo social robot detection. The present invention obtains user data of Sina Weibo through a web crawler, analyzes the behavior of users and robots, and extracts classification features of user data from four different angles, using these features to realize more comprehensive and accurate detection of social robots and the present invention combines the residual neural network, the bidirectional gated recurrent unit and the attention mechanism to build a deep neural network model, which can significantly improve the accuracy and stability of the detection model.

Description

technical field [0001] The invention belongs to the technical field of social robot detection, and in particular relates to a method for detecting a microblog social robot based on a deep neural network. Background technique [0002] In recent years, with the rapid development of information technology, online social networks (OSNs) appear in people's daily life. As OSNs are widely used by people, a social robot account of abnormal users appears. At first, these social robots were used to serve human beings to improve people's quality of life. However, these social robots have begun to be used for some bad or even illegal purposes, so that they have a great negative impact on the entire OSNs and people's daily life. . [0003] In existing OSNs, social robots are a class of special-purpose accounts controlled by programs. They are often used for illegal destruction or profit, including publishing a large amount of advertising information, disseminating malicious URL links, ...

AI Technical Summary

Problems solved by technology

Many research works only focus on extracting the characteristics of a certain aspect of social robots, but cannot achieve a comprehensive description of social robots.

At the same time, even if there are works that extract the features of social robots from multiple perspectives, there are still problems with a small number of features, and it is difficult to fully describe the problem of social robots

[0007] The second problem is that traditional detection models are difficult to effectively solve the problem of social robot detection in the current social network environment

Most of the previous studies have used relatively simple machine learning models to detect social robots in Sina Weibo. Although some results have been achieved, there is still a lot of room for improvement.

Images