Construction Of Social Governance Evaluation Model Based On Sentiment Analysis Of Network Public Opinion

The level of social governance utilizes social indicators to comprehensively reflect social reality and public opinion,grasp the government’s social governance effect,reflect social problems and formulate countermeasures.The traditional social governance index is a comprehensive index based on the official data of government statistics department and questionnaire survey,which is evaluated from six dimensions of human development,social security,social equity,social participation,public security and public service.With the development of research on social governance indicators,social indicators have begun to change from an objective indicator system to a subjective indicator system.The subjective social indicator system is mainly manifested in the degree of satisfaction of individual people with material,spiritual,and cultural aspects.This kind of satisfaction lies in the opinions and opinions expressed by netizens on an event through various social media platforms.The emotional orientation of public opinion on social media networks is analyzed in this paper,measure the satisfaction of the people with the current material,spiritual,and cultural social development,and evaluate the level of social governance in the region from subjective indicators.The main research of this article includes:(1)A Chinese event detection model based on multi granularity context semantic fusion is proposed.In order to solve the problem of mismatch between words and event triggers caused by Chinese word segmentation error.This paper uses the method of character sequence annotation to label the data,and then obtains the semantic information in different word segmentation results through the mechanism of character word fusion gate,so as to solve the problem of mismatch between Chinese words and event triggers.(2)A method for calculating the comprehensive emotional intensity of online public opinion events is proposed.Through the analysis of the emotional tendency of social media texts and comments,combined with the semantic information of punctuation marks and emojis in the text,the emotional strength of the text is obtained;according to the first law of emotional strength,the number of likes of integrated blog posts and comments related to the fusion event,Clicks and other information to get the emotional intensity of the event.Combine the emotional intensity of the text related to the public opinion event with the emotional intensity of the event to obtain the comprehensive emotional intensity of the final network public opinion event.(3)Construct a social governance evaluation model based on sentiment analysis of online public opinion.First,the development status of the event is obtained according to the event trigger word type detection and argument identification.The event development status is used to calculate the event’s impact factor.The event impact factor represents whether the event has a positive or negative impact on the evaluation of social governance.Then multiply the event’s impact factor and the event’s comprehensive emotional intensity to get the event’s social governance impact index.Finally,the social governance impact scores of all events in a region are accumulated and averaged to obtain the social governance index of the region.