Music comment generation method based on deep learning

Abstract

A music comment generation method based on deep learning comprises the following steps: acquiring original music comments by using a web crawler technology, and constructing a music comment library; constructing a text abstract generation model based on the Bert model and the Transformer model, and inputting the lyric text of the to-be-commented music into the text abstract generation model to obtain a lyric abstract text; judging whether a music name or a singer name of the to-be-commented music exists in the original comments of all the music in the music comment library or not; if yes, selecting out the original comments with the music names or the singer names, and if not, calculating the text similarity between the lyric abstract text of the music to be commented and the original comments of each piece of music in the music comment library, and selecting out the original comment with the highest similarity; and constructing a replication generation model based on the VAE model, and inputting the selected original comments into the replication generation model to generate a new comment text. The invention belongs to the technical field of information, and can automatically generate a comment text of music and improve the accuracy, diversity and fluency of the text.

Description

technical field [0001] The invention relates to a method for generating music comments based on deep learning, which belongs to the field of information technology. Background technique [0002] With the development of network technology, users can enjoy all kinds of music anytime and anywhere. In the prior art, there are only a small amount of comments or even no comments in the comment area of ​​a large amount of music. Usually, in order to attract users' participation and attention, a manual method of writing comments can be adopted, but this manual method is very inefficient. [0003] Patent application CN201710196125.X (application name: artificial intelligence-based comment generation method and device, device, and readable medium, application date: 2017.03.29, applicant: Beijing Baidu Netcom Technology Co., Ltd.) discloses an artificial intelligence-based Comment generation method and device, device and readable medium. The method includes: mining candidate word se...

AI Technical Summary

Problems solved by technology

This technical solution only selects replaceable keywords in the target sentence from the corpus, selects the best candidate words for replacement through the scoring system, uses insufficient semantic information of the text, and does not make good use of the overall semantic features of the text and contextual information features, at the same time, the replaced text has a high similarity with the original text, and the accuracy and diversity of the final generated text are also insufficient

Images