Explainable Rec

---

Implementation authors: Junjie Lai and Yue Wu.

This repository is an official PyTorch implementation of the TKDE 2023 paper ExplainableRec Knowledge Enhanced Graph Neural Networks for Explainable Recommendation.

Required packages

The code has been tested running under Python 3.6, with the following packages installed (along with their dependencies):

• pytorch >= 1.0
• numpy >= 1.14.5
• nltk >= 3.5
• gensim >= 3.8.3
• tensorboardX
• wordfreq
• datasketch

Files in the folder

• dataset/

  • Home_and_Kitchen_5.json: a dataset from Amazon 5-core. The dataset in original page (http://jmcauley.ucsd.edu/data/amazon) is inaccessible now, you can download it from Home_and_Kitchen_5.json and put it in there.
  • utils.py: some preprocess methods and tools.
  • text2url.py: which normalizes natural-language text into the ConceptNet URI representation, copy from conceptnet-numberbatch repository.
  • DataManager.py: the script of Dataset and DataLoader.
  • build_dict.py: create the dict from dataset and provide tokenizor

• saved/

  • conceptnet/: you need to download numberbatch-en.txt from conceptnet-numberbatch repository and put it in there.

• modules/: implementations of ExplainableRec.

Training the code

$ cd ExplainableRec
$ python train.py --run_type train --data_path dataset/Home_and_Kitchen_5.json

Testing the code

$ cd ExplainableRec
$ python train.py --run_type test --data_path dataset/Home_and_Kitchen_5.json

Citing ExplainableRec

If you use our ExpalinableRec in your research, please cite:

@ARTICLE{9681226,
author={Lyu, Ziyu and Wu, Yue and Lai, Junjie and Yang, Min and Li, Chengming and Zhou, Wei},
journal={IEEE Transactions on Knowledge and Data Engineering},
title={Knowledge Enhanced Graph Neural Networks for Explainable Recommendation},
year={2023},
volume={35},
number={5},
pages={4954-4968},
doi={10.1109/TKDE.2022.3142260}}