This is our implementation for the paper:
Shaoyun Shi, Min Zhang, Yiqun Liu, and Shaoping Ma. 2018. Attention-based Adaptive Model to Unify Warm and Cold Starts Recommendation. In CIKM'18.
Please cite our paper if you use our codes. Thanks!
Author: Shaoyun Shi (shisy13 AT gmail.com)
@inproceedings{shi2018attention,
title={Attention-based Adaptive Model to Unify Warm and Cold Starts Recommendation},
author={Shi, Shaoyun and Zhang, Min and Liu, Yiqun and Ma, Shaoping},
booktitle={Proceedings of the 27th ACM International Conference on Information and Knowledge Management},
pages={127--136},
year={2018},
organization={ACM}
}
Python 3.5.2
Packages: See in requirements.txt
tensorflow_gpu==1.4.0
pandas==0.23.1
numpy==1.14.5
tqdm==4.23.4
- ml-100k: The origin dataset can be found here. The processed ml-100k dataset is in ./dataset. The codes for processing the data are in ./src/ml-100k.py.
> cd ACCM
> mkdir model
> cd src
# ACCM with Cold-Sampling
> python CSACCM.py --warm_ratio 0.9
# ACCM without Cold-Sampling
> python CSACCM.py --warm_ratio 1.0
Note that other codes ending with
*Model.pyare inherited byCSACCM.py
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent