Microblog event authenticity detection method based on man-machine cooperation

Abstract

The invention discloses a microblog event authenticity detection method based on man-machine cooperation, and the method comprises the following steps: firstly, judging the credibility of a user, and carrying out the static description processing, data reduction and data conversion processing of the user; secondly, detecting emotion scores between the microblogs and the comments and between the microblogs; modeling the event to form a social media network graph model, and processing some data influencing the result manually in the modeling process; and finally, processing and classifying the events by using GCN to obtain a final result which is a classification result of the events. According to the invention, the judgment rule is set for the proposed word level, the emotional tendency of the text is evaluated, and the accuracy of judging whether the microblog event is true or false is improved.

Description

technical field [0001] The invention belongs to the technical field of machine learning, and in particular relates to a method for detecting the authenticity of microblog events. Background technique [0002] A microblog event refers to a microblog news event composed of multiple users expressing their opinions on a popular social news. For a long time, the automatic detection of the authenticity of microblog events has been a research hotspot, and a large number of scholars have used existing technologies to explore the authenticity of microblog events. For example, Zhang, L, Wang, S, Liu, B. Deep learning forsentiment analysis: A survey. WIREs Data Mining Knowl Discov. 2018. Summarized the research methods for sentiment analysis using deep learning, and divided sentiment analysis into three levels : Document level, sentence level and position level, and pointed out the advantages and disadvantages of each level. Document-level sentiment analysis involves long textual dep...

AI Technical Summary

Problems solved by technology

Using GCN for text classification is a good application of GCN, "Yao, Liang&Mao, Chengsheng&Luo, Yuan.Graph Convolutional Networks for Text Classification.(2018)." Constructed a graph on the entire corpus, using a sliding window to obtain Input the matrix, use two layers of GCN to calculate the probability of text category, and convert the true and false detection of events into the classification problem of graph model nodes, but this paper ignores the interaction between texts, so the network structure of the entire event needs to be constructed. mold

Images