The source code for our SIGIR 2022 Paper "Decoupled Side Information Fusion for Sequential Recommendation"
We propose DIF-SR to effectively fuse side information for SR via moving side information from input to the attention layer, motivated by the observation that early integration of side information and item id in the input stage limits the representation power of attention matrices and flexibility of learning gradient. Specifically, we present a novel decoupled side information fusion attention mechanism, which allows higher rank attention matrices and adaptive gradient and thus enhances the learning of item representation. Auxiliary attribute predictors are also utilized upon the final representation in a multi-task training scheme to promote the interaction of side information and item representation.
Our code is based on PyTorch 1.8.1 and runnable for both windows and ubuntu server. Required python packages:
- numpy==1.20.3
- scipy==1.6.3
- torch==1.8.1
- tensorboard==2.7.0
Download datasets from RecSysDatasets or their Google Drive. And put the files in ./dataset/ like the following.
$ tree
.
├── Amazon_Beauty
│ ├── Amazon_Beauty.inter
│ └── Amazon_Beauty.item
├── Amazon_Toys_and_Games
│ ├── Amazon_Toys_and_Games.inter
│ └── Amazon_Toys_and_Games.item
├── Amazon_Sports_and_Outdoors
│ ├── Amazon_Sports_and_Outdoors.inter
│ └── Amazon_Sports_and_Outdoors.item
└── yelp
├── README.md
├── yelp.inter
├── yelp.item
└── yelp.user
Run DIF.sh.
See benchmarks folder to reproduce the results. For example, we show the detailed reproduce steps for the results of DIF-SR on the Amazon Beauty dataset in DIF_Amazon_Beauty.md file.
Due to some stochastic factors, slightly tuning the hyper-parameters using grid search is necessary if you want to reproduce the performance. If you have any question, please issue the project or email us and we will reply you soon.
If you find this repo useful, please cite
@inproceedings{Xie2022DIF,
author = {Yueqi Xie and
Peilin Zhou and
Sunghun Kim},
title = {Decoupled Side Information Fusion for Sequential Recommendation},
booktitle= {International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR)},
year = {2022}
}
This repo is based on RecBole.
Feel free to contact us if there is any question. (YueqiXIE, yxieay@connect.ust.hk; Peilin Zhou, zhoupalin@gmail.com; Russell KIM, russellkim@upstage.ai)