MVIN: Learning Multiview Items for Recommendation, SIGIR 2020

This repository is the implementation of MVIN (arXiv):

Chang-You Tai, Meng-Ru Wu, Yun-Wei Chu, Shao-Yu Chu, and Lun-Wei Ku. SIGIR 2020. MVIN: Learning Multiview Items for Recommendation

We propose the multi-view item network (MVIN), a GNN-based recommendation model which provides superior recommend-ations bydescribing items from a unique mixed view from user and entity angles.

If you want to use our codes and datasets in your research, please cite:

@inproceedings{10.1145/3397271.3401126,
author = {Tai, Chang-You and Wu, Meng-Ru and Chu, Yun-Wei and Chu, Shao-Yu and Ku, Lun-Wei},
title = {MVIN: Learning Multiview Items for Recommendation},
year = {2020},
isbn = {9781450380164},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3397271.3401126},
doi = {10.1145/3397271.3401126},
booktitle = {Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval},
pages = {99–108},
numpages = {10},
keywords = {recommendation, knowledge graph, higher-order connectivity, graph neural network, embedding propagation},
location = {Virtual Event, China},
series = {SIGIR ’20}
}

• data/: datasets
  • MovieLens-1M/
  • amazon-book_20core/
  • last-fm_50core/
• src/model/: implementation of MVIN.
• output/: storing log files
• misc/: storing users being evaluating, popular items, and sharing embeddings.

The code has been tested running under Python 3.6.5. The required packages are as follows:

• tensorflow == 1.12.0
• numpy == 1.15.4
• scipy == 1.1.0
• sklearn == 0.20.0

$ cd MVIN
$ conda deactivate
$ conda env create -f requirements.yml
$ conda activate MVIN

• MVIN

$ cd src/bash/
$ bash main_run.sh "MVIN" $dataset $gpu

• other baseline models

$ cd src/bash/
$ bash main_run.sh $model $dataset $gpu

• some arguments:

• model

  • It specifies the type of model.
  • Here we provide four options, including MVIN and five baseline models:
    • MVIN (by default), proposed in MVIN: Learning Multiview Items for Recommendation, SIGIR 2020. Usage: model=MVIN.
    • KGAT, proposed in KGAT: Knowledge Graph Attention Network for Recommendation, KDD 2019. Usage: model=KGAT.
    • KGCN, proposed in Knowledge Graph Convolutional Networks for Recommender Systems, WWW 2019. Usage: model=KGCN.
    • RippleNet, proposed in RippleNet: Propagating User Preferences on the Knowledge Graph for Recommender Systems, CIKM 2018. Usage: model=RippleNet.
  • You can find other baselines in Github.

• dataset

  • It specifies the dataset.
  • Here we provide three options, including * amazon-book, movie, or last_fm.

• gpu

  • It specifies the gpu, e.g. * 0, 1, and 2.

$ cd src/bash/
$ bash main_att_case_st.sh $gpu

• main_run.sh syntax error near unexpected token elif

$ sed -i -e 's/\r$//' *.sh