2022-KGAT

Reimplementation of the KGAT with Pytorch and DGL(Deep Graph Library)

Knowledge Graph Attention Network

Xiang Wang, Xiangnan He, Yixin Cao, Meng Liu and Tat-Seng Chua (2019). KGAT: Knowledge Graph Attention Network for Recommendation. Paper in ACM DL or Paper in arXiv. In KDD'19, Anchorage, Alaska, USA, August 4-8, 2019.

You can find Tensorflow implementation by the paper authors here and other pytorch version here

Introduction

Knowledge Graph Attention Network (KGAT) is a study recommended the introduction of task knowledge profiles and user behavior data, to explore the higher-order information as side information, thereby enhancing the user to predict based on the user's interaction with an item A question of preference to provide more accurate, more diverse and easier to explain recommendations.

Citation

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

@inproceedings{KGAT19,
  author    = {Xiang Wang and
               Xiangnan He and
               Yixin Cao and
               Meng Liu and
               Tat{-}Seng Chua},
  title     = {{KGAT:} Knowledge Graph Attention Network for Recommendation},
  booktitle = {{KDD}},
  pages     = {950--958},
  year      = {2019}
}

Environment Requirement

The code has been tested running under Python 3.8.10. The required packages are as follows:

CUDA == 11.3
torch == 1.11.0
numpy == 1.22.4
pandas == 1.4.2
scipy == 1.8.1
tqdm == 4.64.0
scikit-learn == 1.1.1
dgl-cu113 = 0.8.1

Run the Codes

From source code

python Main.py --data_name last-fm

From dockerfile

docker pull nvidia/cuda:11.3.0-runtime-ubuntu20.04
# the same dir with dockerfile
docker build -t cuda_2022_kgat:demo .
docker run -it --gpus all --name KGAT_container cuda_2022_kgat:demo

Dataset

We provide three processed datasets: Amazon-book, Last-FM, and Yelp2018.

You can find the full version of recommendation datasets via Amazon-book, Last-FM, and Yelp2018.
We follow KB4Rec to preprocess Amazon-book and Last-FM datasets, mapping items into Freebase entities via title matching if there is a mapping available.

		Amazon-book	Last-FM	Yelp2018
User-Item Interaction	#Users	70,679	23,566	45,919
	#Items	24,915	48,123	45,538
	#Interactions	847,733	3,034,796	1,185,068
Knowledge Graph	#Entities	88,572	58,266	90,961
	#Relations	39	9	42
	#Triplets	2,557,746	464,567	1,853,704

train.txt
- Train file.
- Each line is a user with her/his positive interactions with items: (userID and a list of itemID).
test.txt
- Test file (positive instances).
- Each line is a user with her/his positive interactions with items: (userID and a list of itemID).
- Note that here we treat all unobserved interactions as the negative instances when reporting performance.
final_kg.txt
- Knowledge graph file.
- Each line is a triplet (head_id, relation_id, tail_id) for two entity and one relation in knowledge graph, where head_id and tail_id represent the ID of such entities with relationship relation_id.

Acknowledgement

Any scientific publications that use this datasets should cite the following paper as the reference:

@inproceedings{KGAT19,
  author    = {Xiang Wang and
               Xiangnan He and
               Yixin Cao and
               Meng Liu and
               Tat-Seng Chua},
  title     = {KGAT: Knowledge Graph Attention Network for Recommendation},
  booktitle = {{KDD}},
  year      = {2019}
}

joshua-shuhan / 2022-KGAT-paperreview