# Enhancing Music Features by Knowledge Transfer from User-item Log Data

## TL;DR

This paper introduces a method that leverages user listening logs to enhance music feature extraction by transferring knowledge from user-item interaction data to content-based models, improving performance on classification and regression tasks.

## Contribution

It presents a novel cross-domain knowledge transfer approach that utilizes log data for music feature enhancement, which was previously underexplored for content-based applications.

## Key findings

- Improved accuracy in music classification tasks
- Enhanced regression model performance
- Effective knowledge transfer from log data to content models

## Abstract

In this paper, we propose a novel method that exploits music listening log data for general-purpose music feature extraction. Despite the wealth of information available in the log data of user-item interactions, it has been mostly used for collaborative filtering to find similar items or users and was not fully investigated for content-based music applications. We resolve this problem by extending intra-domain knowledge distillation to cross-domain: i.e., by transferring knowledge obtained from the user-item domain to the music content domain. The proposed system first trains the model that estimates log information from the audio contents; then it uses the model to improve other task-specific models. The experiments on various music classification and regression tasks show that the proposed method successfully improves the performances of the task-specific models.