/CLSR

The official implementation of "Disentangling Long and Short-Term Interests for Recommendation" (WWW '22)

Primary LanguagePythonMIT LicenseMIT

CLSR: Disentangling Long and Short-Term Interests for Recommendation

This is the official implementation of our WWW'22 paper:

Yu Zheng, Chen Gao, Jianxin Chang, Yanan Niu, Yang Song, Depeng Jin, Yong Li, Disentangling Long and Short-Term Interests for Recommendation, In Proceedings of the Web Conference 2022.

The code is tested under a Linux desktop with TensorFlow 1.15.2 and Python 3.6.8.

Please cite our paper if you use this repository.

@inproceedings{zheng2022disentangling,
  title={Disentangling Long and Short-Term Interests for Recommendation},
  author={Zheng, Yu and Gao, Chen and Chang, Jianxin and Niu, Yanan and Song, Yang and Jin, Depeng and Li, Yong},
  booktitle={Proceedings of the ACM Web Conference 2022},
  pages={2256--2267},
  year={2022}
}

Data Pre-processing

Run the script reco_utils/dataset/sequential_reviews.py to generate the data for training and evaluation.

Details of the data are available at Data.

Model Training

Use the following commands to train a CLSR model on Taobao dataset:

cd ./examples/00_quick_start/
python sequential.py --dataset taobao

or on Kuaishou dataset:

cd ./examples/00_quick_start/
python sequential.py --dataset kuaishou

Pretrained Model Evaluation

We provide a pretrained model for the Taobao dataset at Model.

cd ./examples/00_quick_start/
python sequential.py --dataset taobao --only_test

The performance of the provided pretrained model is as follows:

AUC GAUC MRR NDCG@2
0.8954 0.8936 0.4384 0.3807

Note

The implemention is based on Microsoft Recommender.