Research On Rumors Detection Method Of Corona Virus Disease 2019 Based On Text Classification

In recent years,Internet-related technologies have been developed rapidly,and self-media platforms have sprung up rapidly.The self-media platform has also become an indispensable part of the lives of netizens.The number of users has grown exponentially.With the outbreak of the COVID-19,the information disseminated by the self-media platform has become a huge epidemic database,but with it What’s wrong is that all kinds of false information are mixed in the Internet and spread freely among the crowd,had caused great distress to people and seriously affected the security and stability of society.Automating rumor identification on rumor information on social media sites has become an important issue.This article aims to clean and label the news text in news texts without structural rules to obtain structured data and construct a new crown epidemic rumors data set.Through the feature extraction of the data set,a user model and a variable-length time series model are constructed,and based on this,it is judged whether the "rumor" text is a rumor.This paper also combines the text classification method with the spread mode of rumors to try to get a greater breakthrough in the task of rumor detection.The main work of this paper is to use deep learning methods to mine hidden features from a large number of “rumors” texts of the COVID-19,and establish a time series model that can not only use the content information of the text,but also make full use of user information.The dissemination mode of the text,the main contributions of this paper are as follows:(1)Deeply excavate the blog post text information and user comment information,establish a user comment model,and use the user’s attitude to the news as an important basis for detecting whether the news text is a rumor.(2)Propose a variable-length time series model,extract topics from the existing rumor data set,and generate time stamps based on the time series labeling method,and then extract data based on the blog post text content,user characteristics,and communication characteristics,and finally verify that it can be The utility of variable length time series model for rumor detection.(3)The design of a human-machine collaborative rumor detection system is used to assist individuals in the collection of rumor texts,manual data annotation,and automatic annotation of texts by embedding rumor detection algorithms.