Singer classification method based on Labeled LDA model

Abstract

The invention relates to a singer classification method based on a Labeled LDA model. The singer classification method comprises the steps of S1 collecting and preprocessing artificial tags of singers; S2 establishing a singer classification model based on user behaviors and collecting user behavior data; S3 cleaning the user behavior data, and filtering data unfavorable for model training in theuser behavior data; S4 distributing the weight of each singer corresponding to each user in the user behavior data; S5 combining the user behavior data with the artificial label data to generate training data; and S6 based on the training data, referring to the label combination relationship, and carrying out Labeled LDA model training based on optimized Gibbs sampling. According to the invention,the song playing behavior of the user is used as training data, the coverage of the user is high, the preference characteristics of each user group are considered, the change of the user behavior reflects the change of social hotspots and public cognition, the model can be periodically trained to change along with the change, the adaptability is strong, the precision degree is high, the label coverage rate is improved, and the classification is fine enough.

Description

technical field [0001] The present invention relates to the technical field of Internet individualized service, specifically a kind of singer classification method based on Labeled LDA model. Background technique [0002] In the past ten years, the rapid development of Internet music has gradually eroded the traditional music market. Internet music manufacturers such as Tencent Music Group, Netease Cloud Music, and Xiami Music have entered thousands of households. In the traditional music market, in addition to limited promotion methods such as TV, movies, and the Internet, users generally learn about new music (songs) in record stores. [0003] Internet music uses the music app as the channel. Users face an unprecedented variety of music choices on the music app, but it is impossible for users to know every song and every singer. Therefore, an effective information filtering method is needed to help users filter songs. . Nowadays, the number of Internet users is generall...

AI Technical Summary

Problems solved by technology

After the model extracts labels for singers, it still cannot evaluate the genre attributes of singers or music from an intuitive perspective, and it is poorly adaptable to some application scenarios

[0009] (2) The classification of categories is uncontrollable, so the number of categories is uncertain, and the number of items contained in each category is also uncertain

This is very unfriendly to many recommendation algorithms, and may have an impact on the implementation effect, and may also have an impact on the implementation efficiency

[0010] (3) Using the same data for learning, often times the learning results are inconsistent

For most singers who are not so popular or have a vague style, the coverage of artificial tags is very limited

[0013] (2) Many singers will have multiple artificial labels at the same time, and experts cannot assign a weight to each label of the singer

Therefore, the labels given by experts to singers do not necessarily fully conform to the perception of users.

[0015] (4) The labels assigned by experts to singers are generally fixed and cannot change over time, but the genre of singers is likely to change over time from the perspective of users

Images