Exploration Of Moral Emotions In The Online Environment Based On Big Data Technology

Cyberspace is an important communication carrier of today’s public events.On the basis of mutual penetration with real life,it has greatly expanded the channels for citizens to express their opinions and emotions,so that opinions and emotions have gradually become an important part of hot events.However,it should not be ignored that the comments on the network platform with rich moral emotions can get rid of the constraints of space and have a significant impact on the real society.Therefore,a systematic understanding and in-depth exploration of the moral emotions behind the current online comments is conducive to the improvement of individual sense of wellbeing of the public and the maintenance of online social morality.It provides support for correctly guiding moral emotions and maintaining and promoting public order.With the aid of big data technology,the present study built a moral sentiment classification tool to conduct a quantitative analysis of moral emotions in the network environment and provide tool support for the subsequent research on moral emotions.At the same time,taking the hot issues into account,we explored the characteristics of moral emotions in the network environment in order to provide theoretical basis and practical guidance for the maintenance of cyberspace moral construction.In order to quantify the moral emotion of online comments,the moral emotion dictionary and the moral emotion text classification model were constructed in study 1,and the tools were evaluated.In terms of moral emotion dictionary,this paper used the widely used emotion dictionary of Dalian Institute of Technology as the basic dictionary,and used Word2 vec technology to select similar words to form the basic moral emotion dictionary.At the same time,auxiliary dictionaries of degree words,negative words and stop words were built,and the corresponding dictionary calculation rules were formulated.We applied deep learning technology to moral emotion text classification model.On the basis of the ERNIE-TINY model,we implemented the transfer learning of moral emotion classification model.Finally,the classification effect of the above two methods is evaluated and selected according to the annotation set.In order to further explore the characteristics of moral emotions in network events,the second study classified the moral emotions of the blog posts related to "Zhang Guimei" and "Wu Yifan" with the help of the tool of moral emotions classification built in study 1,analyzed the event development and time characteristics,and drew the following conclusions: Due to the virtualization of cyberspace,the diversification of information fragments,the substitution of private morality for public morality and other factors,the network moral emotions are characterized by diversification,less reflective moral emotion,and the enhancement of diffusion effect.Accordingly,this study puts forward some suggestions from the aspects of legal system construction,moral atmosphere construction,and network moral cultivation.Generally speaking,this study proved the validity of moral emotion dictionary and moral emotion classification model and the feasibility of the method through the quantitative analysis of moral emotion of network events,and discussed the characteristics of moral emotion in cyberspace,providing some reference for the follow-up moral construction in cyberspace.