/CPFairRecSys

[Official Codes] CPFair: Personalized Consumer and Producer Fairness Re-ranking for Recommender Systems (SIGIR2022)

Primary LanguageJupyter NotebookMIT LicenseMIT

CPFairRecSys

The re-ranking method for a fair recommendation w.r.t both users and items dimension.

Some Statistics

some statistics of the papers

Dataset Prepration

  • ratings_data.txt: user-item interactions (raw file) (by download, from Cornac or another resource)
  • [DatasetName]_data.txt: This file is a k-core file that the users and items ids are mapped to a new range of indecies.

Datasets

All the datasets used in the experiments are in the datasets folder. Each dataset contains several files and a folder including:

  • Folder:
    • groups: includes two sub-folders, items and users, in the items folder we have the item groups (longtail_items.txt and shorthead_items.txt) and in users folder we have two folders (005 and 020) each includes two user group files, i.e., active_ids.txt and inactive_ids.txt.
  • Files:
    • ratings_data.txt
    • [DatasetName]_data.txt
    • [DatasetName]_inters.txt
    • [DatasetName]_train.txt
    • [DatasetName]_test.txt
    • [DatasetName]_tune.txt

Model

To run them model you need to run the fair_model notebook:

https://colab.research.google.com/github/rahmanidashti/CPFairRecSys/blob/main/fair_model.ipynb

Team

Mohammadmehdi Naghiaei, DECIDE, University of Southern California

Hossein A. Rahmani, Web Intelligence Group, UCL

Yashar Deldjoo, SisInf Lab, Polytechnic University of Bari

Citation

If you use our source code, dataset, and experiments for your research or development, please cite the following paper:

@inproceedings{naghiaei2022cpfairness,
  title={CPFair: Personalized Consumer and Producer Fairness Re-ranking for Recommender Systems},
  author={Mohammadmehdi Naghiaei, Hossein A. Rahmani, Yashar Deldjoo},
  booktitle={The 45th International ACM SIGIR Conference on Research and Development in Information Retrieval},
  year={2022}
}
@article{Naghiaei2022PyCPFair,
title = {PyCPFair: A framework for consumer and producer fairness in recommender systems},
journal = {Software Impacts},
pages = {100382},
year = {2022},
issn = {2665-9638},
doi = {https://doi.org/10.1016/j.simpa.2022.100382},
url = {https://www.sciencedirect.com/science/article/pii/S2665963822000835},
author = {Mohammadmehdi Naghiaei and Hossein A. Rahmani and Yashar Deldjoo}
}

Contact

If you have any questions, do not hesitate to contact us by h.rahmani@ucl.ac.uk or rahmanidashti@gmail.com, we will be happy to assist.