Research On Users’ Emotional And Interactive Behaviors In Online Comments Of Government Microblog

The booming online services of social media such as microblog have attracted a tremendous number of users to participate.Government microblog has become an important source of information for the public to focus on social problems and perceive political situations.It is imperative that the government should promote and encourage the rapid development of the online social media industry,and regulate and restrict the behaviors on online social media.China has always attached great importance to the ideological attributes as well as publicity and educational functions of the media.Therefore,the government and the society are supposed to consider how to accurately grasp the development trend of the Internet,make full use of information science and technology,pay attention to the emotional tendencies and interactive behaviors of online social media users,and thus boost the great and rapid development of online social media.The considerable amount of data generated by social media users on social networks has brought a wealth of research materials to the academic circles.Although there have been some meaningful explorations in this hot field,they are still insufficient.This paper is mainly based on the current research on online social media,driven by theories such as Affective Response Theory and Social Network Theory,using empirical analysis and machine learning methods to study users’ emotional expressions and interactive behaviors in comments of government microblog.Specifically,there are three parts:1.To investigate the generation mechanism of emotional tendencies of government microblog users.Based on the theory of Affective Response,combined with Media Richness Theory,Emotional Contagion Theory,Theory of Reasoned Action,etc.,a model was constructed to study the influential factors of users’ emotional expressions in the comments of government microblog,and to explore the role of gender.The results show that the more informative and positive the government microblog information is,the more positive comments it will receive from users;users who are more socially active and have more positive emotional habits in historical posts are more likely to give positive comments on other tweets;female users are more susceptible to information,while male users are more likely to express emotions rationally based on their traits.In order to enhance users’ positive comments,microblog service providers can refer to the research results and implement a gender difference strategy.For example,they can provide female users with richer information;for male users,they can combine the contents with their historical interests.The government can also benefit from learning the mechanism of online emotional expressions of users,such as using more topic tags to guide users’ emotions.2.To investigate the generation mechanism of interactive behaviors of government microblog users.Based on Social Network Theory,combined with Social Learning Theory,Homogeneity Theory,Social Capital Theory,etc.,an ERGM research model was constructed to explore the factors of the formation of the interaction network among users in government microblog.The research confirms the impact of reciprocity,transitivity,homogeneity,social capital,and emotion on the formation of interactive networks among users represented by comments/responses.According to the conclusions,microblog service providers and the management departments can take appropriate actions or policies to guide the formation of interaction networks among users in government social media,such as attracting more common users to participate with users who have more social influence or higher social activeness,stimulating users’ interaction willingness and increasing the participation of the government microblog platform by establishing interest-based user groups,and encouraging more users to discuss by publishing some hot events or setting some special topics(such as controversial social phenomena and social problems).3.To predict user interaction behavior and emotional tendency in government microblog.Based on machine learning,the paper realized the prediction of user interaction behaviors and user emotional tendencies represented by comments/responses.The study found that using the method of MLP with Doc2 vec word vector features of microblog information and user historical posts,microblog information feature groups,and user feature groups can predict user interactional behaviors represented by comments/responses and user emotional tendencies.In addition,the validity of the neural network model is verified through comparative experiments.When microblog service providers,enterprise managers,and public sector managers operate online social media platforms,they can refer to the research conclusions to predict user emotional tendencies and interactive enthusiasm for tweets published by the platform,thereby improving management and service efficiency,which is of great significance for the formulation and implementation of public policies and is crucial for the government to manage and optimize online social media.