quxiaolong0812 / mkgcn

Multi-modal Knowledge Graph Convolutional Networks for Music Recommendation System

MKGCN4M

This repository is the implementation of the MKGCN:

MKGCN: Multi-modal Knowledge Graph Convolutional Networks for Music Recommender System

Designed by Qu Xiaolong in Beijing Forestry University.

Files in the folder

• data/
  • music-d/
    • itemID2entityID.txt: the mapping from item indices in the raw rating file to entity IDs in the KG;
    • kg.txt: knowledge graph file;
    • userID2itemID4ratings.csv: raw rating file of music-d;
    • modals/: 7 modals files for music-d dataset;
    • data_config.py: dataset configuration file;
• other .py files: implementations of MKGCN.

Attention: music-d is multi-modal music dataset by myself. Actually, I have created four multimodal music datasets of different sizes, but due to the need for data copyright and confidentiality, I only open the smallest one.

Multimodal files definition are shown as below:

Required packages

The code has been tested running under Python 3.6.5, with the following packages installed (along with their dependencies):

• pytorch == 2.0.0
• numpy == 1.14.5
• sklearn == 0.24.2

Running the code

$ python main.py --dataset music-d (note: use -h to check optional arguments)

Note

I will update more information in the future, please stay tuned.

mkgcn's Issues

数据集

您好！
在细读您的研究论文后，我深受启发，对您的模型产生了浓厚的兴趣。我希望能够基于您的模型，复现论文中的成果，以便进一步的学术研究和探索。因此，我想咨询您，是否有可能分享除已提供数据集之外的任意一个其他的数据集呢？
非常感谢您的帮助！