| With the rapid development of online video industry in China,it has become an important concern of video managers to promote user activity and retention and improve video popularity.Bullet screens and comments are the main ways for users to express their ideas and communicate with other users,which has a significant impact on the popularity of videos.Although both of these forms contain rich emotional information,there are obvious differences in language characteristics and immediacy of information,which have different impacts on users.At present,scholars have only preliminarily explored the differences between bullet screens and comments in terms of emotional inclination and interactivity,but there is a lack of in-depth research on their impact on video popularity.Therefore,this thesis takes Station B as the research object,conducts sentiment analysis and theme mining on the video barrage and comment,and explores the influence of the video barrage and comment on the video popularity and their contribution differences in the videos with both barrage and comment.First,according to relevant theories,a research model of influencing factors of video popularity based on the emotion analysis of bullet screen and comments is constructed to explore the influence of relevant variables on video popularity and put forward relevant hypotheses.Taking Station B as the research object,tracking and crawling sample the data of this video for 10 weeks.Then,according to the language characteristics of bullet screen and comments,the appropriate emotion dictionary is constructed,and its emotion value is calculated.Stepwise regression analysis was carried out on the processed data,and the influences of different emotional tendencies of barrage and comments on the indicators measuring the popularity of videos such as the number of plays,the number of coins and the number of favorites were explored through four-time regression,and the regulating effect of the number of video uploaders was analyzed to verify the research hypothesis.Finally,according to the empirical research conclusion,the theme analysis of the bullet screen and comments that can improve the popularity of video emotional tendency is carried out.Through the empirical research on the effect of the emotion of barrage and comment on the popularity of video and the thematic analysis of barrage and comment,the following conclusions can be drawn:(1)The positive emotion of comment has a significant positive impact on the popularity of video,while the negative emotion of comment has no significant impact.(2)The positive emotion of bullet screen has a significant positive influence on the amount of videos played,the amount of coins inserted and the amount of favorites.The negative emotion of bullet screen has a significant positive effect on the amount of videos played and the amount of collections,but has no significant effect on the amount of coins.(3)The number of fans of video uploaders has a negative moderating effect on the impact of negative emotion of bullet screen on video popularity,while the number of fans of video uploaders has a positive moderating effect on the impact of positive emotion of bullet screen on video popularity.(4)Compared with the impact of comment emotion on video popularity,bullet screen emotion has a more significant positive impact on video popularity.(5)Users pay much more attention to a certain topic in bullet screens and comments than other topics.(6)In bullet screen,users pay more attention to broadcast emotional confession content,while in positive comments,they pay more attention to interactive relaxed and happy topics,which is related to the different characteristics of bullet screen and comments.This thesis specifically analyzes the impact of the emotion of bullet screen and comment on the video popularity indicators,and studies the user’s focus on the two interaction modes.Enriched the application of relevant theories on video social platforms,and provided some enlightenment and suggestions for video creators and platforms. |
PDF Full Text Request