Design And Implementation Of Music Introduction Generating System Based On Text Information

With the widespread popularity of the Internet and the development of online music,the mechanism of online music has been continuously improved,and the development of online music platforms has entered a mature stage.Among it,text data such as music introduction and review play a great role in the openness,interactivity,and sharing of online music platforms.It can enable the platform users to obtain the relevant information of the song more conveniently and quickly,and the music can also be communicated and shared on a larger scale.Therefore,the use of natural language generation technology to automatically generate high-quality music introduction text to introduce and recommend the content and characteristics of the music is very promising and practical.According to the characteristics of the music introduction text generation task,this paper analyzes the language characteristics of the target text,and proposes a music introduction text generation method based on end-to-end deep learning model.First,we use the codec model combined with the pre-trained Bert model to generate the lyrics summary of the song,filter the collected natural comment text through text comparison and keyword matching,and then use the paraphrase model to rewrite the text,and combine the variational auto-encoder structure to optimize the diversity and creativity of the generated text.The results generated by the two-part model are combined to obtain a music introduction text composed of two parts:lyrics summary and song analysis.In addition to the research and improvement of the algorithm,this article also uses data collecting and manual screening methods to produce a new data set,and finally obtains a self-made data set of Chinese song information and corresponding comment texts under 72 music categories.The field of this data set fits well with the research topic of this article,and to a certain extent reflects the real situation of current online music text,which can better verify the effectiveness of the generation algorithm.After a large number of comparative experiments on the public data set and the self-made music text data set,compared with the existing models,the results of the proposed algorithm have improved the fluency,accuracy and diversity of the generated text.The experimental results on the self-made music text data set confirm the feasibility of the generation algorithm,which can effectively generate the introduction text of the target song.On the basis of this model,this paper designs and implements a music introduction text intelligent generation system based on Web development.It can complete the online operation process of the text generation function through the service request and data communication between the front-end,back-end and model of the webpage.Finally,this paper deploys the system and conducts functional test and performance test to verify the effectiveness of the system.