HINRL4Rec: a novel integrated heterogeneous network embedding with reinforcement learning for recommendation
This is the source code of HINRL4Rec model which is an integrated rich-textual heterogeneous network embedding with the policy-guided path-based searching mechanism via reinforcement learning approach for explainable recommendation task.
Our works and this implemented source code are majorly inspired and inherited from previous works of Xian, Y. et al. in the proposed PGPR model [1] (https://github.com/orcax/PGPR).
[1] PGPR model:
@inproceedings{xian2019reinforcement,
title={Reinforcement knowledge graph reasoning for explainable recommendation},
author={Xian, Yikun and Fu, Zuohui and Muthukrishnan, S and De Melo, Gerard and Zhang, Yongfeng},
booktitle={Proceedings of the 42nd international ACM SIGIR conference on research and development in information retrieval},
pages={285--294},
year={2019}
}
- Python >= 3.7
- PyTorch >= 1.7.1
- Tensorflow >= 2.3.1
- Keras >= 1.0.8
- MovieLens: https://grouplens.org/datasets/movielens/
- Amazon reviews (Stanford - SNAP): http://snap.stanford.edu/data/web-Amazon-links.html
Please send any question you might have about the code and/or the algorithm to [email protected].