Study On User’s Emotion Of Music Social Platform

With the rapid development of Internet music platforms and the use of emotional marketing in recent years,related fields have begun to realize that Internet enterprise marketing should focus on mining the emotional characteristics of customers,especially for the music social market with particularly distinctive emotional characteristics.The essence of music is emotion.In the emotional communication of “music social”,emotional elements have a more special status and significance.The music social platform can carry out targeted emotional marketing and improve user loyalty by capturing user characteristic information.User interaction comments are a key feature of the music social platform,which implies emotional feedback of a large number of user music experiences.The traditional user information mining is mainly based on the behavior level information such as access records,ignoring the contents of the user’s music emotional psychology implied in the user comments.Whether the user enjoys the music appreciation,whether it is the whole or the individual emotional law Information that is important for marketing activities and platform development.In addition,there are still some shortcomings and many arguments about the theory of imagery in music emotions and musical emotions,which need to be supplemented by practice.Based on the user commentary information on Net Ease cloud music,this paper focuses on the analysis of related issues in the characteristics of image processing in music emotion.Aiming at the lack of emotional marketing in music social communication,the author proposes the“three-way two-degree” evaluation index of user emotion and applies it to the research conclusions.In the research,the Python language was used to realize the crawling of the Nete cloud cloud user user review data,the construction of the subsequent model,and the data visualization.The specific research consists of two parts.The basic research part mainly confirms the reliability of user comments on the classification of music emotion categories.It is the basis for the follow-up research of user comments as data objects.The music works with specific emotional labels are taken as the research object,and the corresponding music reviews are used as the training and test corpus of the sentiment classification model.The trained models are used to conduct statistics and analysis of various indicators such as the accuracy of music emotion classification.Combining the results of music sentiment analysis,the paper discusses and summarizes the usefulness and reliability of music commentary in the study of musical emotion discrimination.The main research part optimizes the traditional LDA topic model as atool for image data mining in this paper by integrating emotion and semantic elements.Based on the basic research,the basic topic distribution under each emotion category and the specific image corresponding to each topic are obtained.Through the word2 vec word vector model trained by the pre-study institute,the subject image is mapped to the 100-dimensional semantic vector space,and the distribution feature values ??of each basic topic are counted.At the same time,the distribution characteristics of the subject image in the semantic space are observed in the two-dimensional space.In the judgment of sentiment orientation,an emotional dictionary suitable for the music review data of this article is established,the sentiment orientation judgment is made to each image and topic,the emotional tendency law of the topic distribution under each emotion category is statistically analyzed,and the corresponding emotional feature evaluation matrix is ??determined.According to the user’s emotional characteristics and its evaluation matrix in the conclusions of this paper,although most of the emotional responses of users in the three types of emotional works are positive positive bias,compared with the emotions of joy,happiness and tranquility,users of music social platforms are sad.Lonely and other low moods have richer semantic content.In the targeted emotional marketing,the music platform should grasp the user’s sentimental cultural needs to promote user retention and improve marketing efficiency.In response to the basic questions raised in the study,the specific conclusions of this study are:1)The user’s music comment can clearly distinguish the music mood;2)Obtain the main topics of various musical emotions,as well as the specific image content under each topic;3)On the social platform,the user shows a positive positive emotional tendency towards all kinds of emotional music works;4)The sad mood of the user’s music experience has a richer and more extensive image content;5)The emotional characteristics of users on the music social platform mainly show a wide range of sentimental cultural topicss;6)Obtained the main connotation of the user’s sentimental cultural needs.