Internet public opinion early warning mechanism analysis method based on RVM-L model

Abstract

The invention discloses an Internet public opinion early warning mechanism analysis method based on an RVM-L model, and the method comprises the steps: 1, collecting a certain type of public opinion information data on the Internet, and carrying out the regression fitting of the public opinion information data at all time intervals through an RVM regression model; 2, inputting an obtained heat function into a Logistic model, and carrying out trend prediction of the function to obtain a Logistic function; solving the Logistic function to obtain a key point of public opinion trend change; 3, according to the key point of public opinion trend transformation, obtaining a rapid public opinion growth interval and a maximum public opinion growth trend; giving a threshold value, and calculating acorresponding early warning critical zone when the Logistic function value is smaller than the threshold value; taking public opinion coping measures in the early warning interval to achieve the effective control over public opinion outbreak. According to the method, the public opinion information influenced by multiple factors can be effectively fitted, the trend is effectively analyzed, the critical zone of the key point is accurately predicted, and the effective early warning function of internet public opinions generated by dealing with emergencies is realized.

Description

technical field [0001] The invention relates to the technical field of social networks, in particular to an analysis method of an Internet public opinion early warning mechanism based on an RVM-L model. Background technique [0002] According to the 44th "Statistical Report on Internet Development in China" released by China Internet Network Information Center in August 2019, as of June 2019, the number of Internet users in my country reached 854 million, the Internet penetration rate reached 61.2%, and the search engine utilization rate was 81.3% %. The rapid increase in the number of users of Internet platforms has made the speed and scale of information gathering much faster than before. Public opinion monitoring has also shifted from public opinion tracking to public opinion early warning. [0003] Researchers have proposed different models for Internet public opinion analysis to improve prediction accuracy, mainly including: time series-based prediction model, which use...

AI Technical Summary

Problems solved by technology

[0004] The technical problem to be solved by the present invention is that in the prior art, in the Internet public opinion, the public opinion information has the characteristics of fast data accumulation, difficult judgment of key points, and many factors affecting information, and it is difficult to improve the prediction effect of the public opinion prediction model. Provide

Images