The repo for "MGIPF: Multi-Granularity Interest Prediction Framework for Personalized Recommendation", SIGIR 2025
conda create -n mgipf python=3.8
conda activate mgipf
pip install -r requirements.txt
- Download and unzip 'UserBehavior.csv.zip' (https://tianchi.aliyun.com/dataset/649) to
data_user/ - Download and unzip 'data_format1.zip' (https://tianchi.aliyun.com/dataset/42) to
data_ijcai/ - Download the 4 files of Taobao (https://tianchi.aliyun.com/dataset/56) and unzip to
data_taobao/raw_data/
- Modify the 'dataset' in
run.sh - Run
run.shand wait
@inproceedings{feng2025mgipf,
title={MGIPF: Multi-Granularity Interest Prediction Framework for Personalized Recommendation},
author={Feng, Ruoxuan and Tian, Zhen and Peng, Qiushi and Mao, Jiaxin and Zhao, Wayne Xin and Hu, Di and Zhang, Changwang},
booktitle={Proceedings of the 48th International ACM SIGIR Conference on Research and Development in Information Retrieval},
pages={2172--2181},
year={2025}
}