Consumer Fairness in Recommender Systems: Contextualizing Definitions and Mitigations

Overview

Consumer Fairness in Recommender Systems: Contextualizing Definitions and Mitigations

reproducibility task

This is the repository for the paper Consumer Fairness in Recommender Systems: Contextualizing Definitions and Mitigation, developed by Giacomo Medda, PhD student at University of Cagliari, with the support of Gianni Fenu, Full Professor at University of Cagliari, Mirko Marras, Non-tenure Track Assistant Professor at University of Cagliari, and Ludovico Boratto, Tenure Track Assistant Professor at University of Cagliari.

The goal of the paper was to find a common understanding and practical benchmarks on how and when each procedure of consumer fairness in recommender systems can be used in comparison to the others.

Repository Organization

  • reproducibility_study

    This is the directory that contains the source code of each reproduced paper identified by the author names of the respective paper.

    • Ashokan and Haas: Fairness metrics and bias mitigation strategies for rating predictions
    • Burke et al: Balanced Neighborhoods for Multi-sided Fairness in Recommendation
    • Ekstrand et al: All The Cool Kids, How Do They Fit In. Popularity and Demographic Biases in Recommender Evaluation and Effectiveness
    • Frisch et al: Co-clustering for fair recommendation
    • Kamishima et al: Recommendation Independence
    • Li et al: User-oriented Fairness in Recommendation
    • Rastegarpanah et al: Fighting Fire with Fire. Using Antidote Data to Improve Polarization and Fairness of Recommender Systems
    • Wu et al: Learning Fair Representations for Recommendation. A Graph-based Perspective
  • Preprocessing

    Contains the scripts to preprocess the raw datasets and to generate the input data for each reproduced paper.

  • Evaluation

    Contains the scripts to load the predictions of each reproduced paper, compute the metrics and generate plots and tables in latex and markdown forms.

  • Other Folders

    The other folders not already mentioned are part of the codebase that supports the scripts contained in Preprocessing and Evaluation. These directories and their contents are described by README_codebase, since the structure and code inside these folders is only used to support the reproducibility study and it is independent from the specific implementation of each paper.

Reproducibility Pipeline

  • Code Integration.

    The preprocessing of the raw datasets is performed by the scripts.

    The commands to preprocess each dataset are present at the top of the related dataset script, but the procedure is better described inside the REPRODUCE.md. The preprocessed datasets will be saved in data/preprocessed_datasets.

    Once the MovieLens 1M and the Last.FM 1K dataset have been processed, we can pass to the generation of the input data for each reproduced paper:

    The commands to generate the input data for each preprocessed dataset and sensitive attribute are present at the top of the script, but the procedure is better described inside the REPRODUCE.md). The generated input data will be saved in Preprocessing/input_data.

  • Mitigation Execution

    Each paper (folder) listed in the subsection reproducibility_study of Repository Organization contains a REPRODUCE.md file that describes everything to setup, prepare and run each reproduced paper. In particular, instructions to install the dependencies are provided, as well as the specific subfolders to fill with the input data generated in the previous step, in order to properly run the experiments of the selected paper. The procedure for each source code is better described in the already mentioned REPRODUCE.md file.

  • Relevance Estimation and Metrics Computation

    The REPRODUCE.md file contained in each "paper" folder describes also where the predictions can be found at the end of the mitigation procedure and guide the developer on following the instructions of the REPRODUCE.md of Evaluation that contains:

    • metrics_reproduced: script that loads all the predictions of relevance scores and computes the metrics in form of plots and latex tables This is the script that must be configured the most, since the paths of the specific predictions of each paper and model could be copied and pasted inside the script if the filenames do not correspond to what we expect and prepare. The REPRODUCE.MD already mentioned better described these steps and specifying which are the commands to execute to get the desired results.

Installation

Considering the codebase and the different versions of libraries used by each paper, multiple Python versions are mandatory to execute properly this code.

The codebase (that is the code not inside reproducibility_study, Preprocessing, Evaluation) needs a Python 3.8 installation and all the necessary dependencies can be installed with the requirements.txt file in the root of the repository with the following command in Windows:

pip install -r requirements.txt

or in Linux:

pip3 install -r requirements.txt

The installation of each reproducible paper is thoroughly described in the REPRODUCE.md that you can find in each paper folder, but every folder contains a requirements.txt file that you can use to install the dependencies in the same way. We recommend to use virtual environments at least for each reproduced paper, since some require specific versions of Python (2, 3, 3.7) and a virtual environment for each paper will maintain a good order in the code organization. Virtual environments can be created in different ways depending on the Python version and on the system. The Python Documentation describes the creation of virtual environments for Python >= 3.5, while the virtualenv Website can be used for Python 2.

Results

Top-N Recommendation Gender

Top-N Recommendation Gender

Top-N Recommendation Age

Top-N Recommendation Age

Rating Prediction Gender

Rating Prediction Gender

Rating Prediction Age

Rating Prediction Age

Learning and Building Convolutional Neural Networks using PyTorch

Image Classification Using Deep Learning Learning and Building Convolutional Neural Networks using PyTorch. Models, selected are based on number of ci

Mayur 126 Dec 22, 2022
Source code for "Roto-translated Local Coordinate Framesfor Interacting Dynamical Systems"

Roto-translated Local Coordinate Frames for Interacting Dynamical Systems Source code for Roto-translated Local Coordinate Frames for Interacting Dyna

Miltiadis Kofinas 19 Nov 27, 2022
A program that can analyze videos according to the weights you select

MaskMonitor A program that can analyze videos according to the weights you select 下載 訓練完的 weight檔案 執行 MaskDetection.py 內部可更改 輸入來源(鏡頭, 影片, 圖片) 以及輸出條件(人

Patrick_star 1 Nov 07, 2021
Scalable Graph Neural Networks for Heterogeneous Graphs

Neighbor Averaging over Relation Subgraphs (NARS) NARS is an algorithm for node classification on heterogeneous graphs, based on scalable neighbor ave

Facebook Research 67 Dec 03, 2022
Using LSTM to detect spoofing attacks in an Air-Ground network

Using LSTM to detect spoofing attacks in an Air-Ground network Specifications IDE: Spider Packages: Tensorflow 2.1.0 Keras NumPy Scikit-learn Matplotl

Tiep M. H. 1 Nov 20, 2021
Gradient representations in ReLU networks as similarity functions

Gradient representations in ReLU networks as similarity functions by Dániel Rácz and Bálint Daróczy. This repo contains the python code related to our

1 Oct 08, 2021
Source code to accompany Defunctland's video "FASTPASS: A Complicated Legacy"

Shapeland Simulator Source code to accompany Defunctland's video "FASTPASS: A Complicated Legacy" Download the video at https://www.youtube.com/watch?

TouringPlans.com 70 Dec 14, 2022
An Evaluation of Generative Adversarial Networks for Collaborative Filtering.

An Evaluation of Generative Adversarial Networks for Collaborative Filtering. This repository was developed by Fernando B. Pérez Maurera. Fernando is

Fernando Benjamín PÉREZ MAURERA 0 Jan 19, 2022
Roger Labbe 13k Dec 29, 2022
RoIAlign & crop_and_resize for PyTorch

RoIAlign for PyTorch This is a PyTorch version of RoIAlign. This implementation is based on crop_and_resize and supports both forward and backward on

Long Chen 530 Jan 07, 2023
Training code and evaluation benchmarks for the "Self-Supervised Policy Adaptation during Deployment" paper.

Self-Supervised Policy Adaptation during Deployment PyTorch implementation of PAD and evaluation benchmarks from Self-Supervised Policy Adaptation dur

Nicklas Hansen 101 Nov 01, 2022
CAMoE + Dual SoftMax Loss (DSL): Improving Video-Text Retrieval by Multi-Stream Corpus Alignment and Dual Softmax Loss

CAMoE + Dual SoftMax Loss (DSL): Improving Video-Text Retrieval by Multi-Stream Corpus Alignment and Dual Softmax Loss This is official implement of "

程星 87 Dec 24, 2022
Header-only library for using Keras models in C++.

frugally-deep Use Keras models in C++ with ease Table of contents Introduction Usage Performance Requirements and Installation FAQ Introduction Would

Tobias Hermann 927 Jan 05, 2023
Official implementation of our neural-network-based fast diffuse room impulse response generator (FAST-RIR)

This is the official implementation of our neural-network-based fast diffuse room impulse response generator (FAST-RIR) for generating room impulse responses (RIRs) for a given acoustic environment.

12 Jan 13, 2022
UA-GEC: Grammatical Error Correction and Fluency Corpus for the Ukrainian Language

UA-GEC: Grammatical Error Correction and Fluency Corpus for the Ukrainian Language This repository contains UA-GEC data and an accompanying Python lib

Grammarly 226 Dec 29, 2022
Official implementation of "Watermarking Images in Self-Supervised Latent-Spaces"

🔍 Watermarking Images in Self-Supervised Latent-Spaces PyTorch implementation and pretrained models for the paper. For details, see Watermarking Imag

Meta Research 32 Dec 13, 2022
Relaxed-machines - explorations in neuro-symbolic differentiable interpreters

Relaxed Machines Explorations in neuro-symbolic differentiable interpreters. Baby steps: inc_stop Libraries JAX Haiku Optax Resources Chapter 3 (∂4: A

Nada Amin 6 Feb 02, 2022
Pytorch Implementation of Google's Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling

Parallel Tacotron2 Pytorch Implementation of Google's Parallel Tacotron 2: A Non-Autoregressive Neural TTS Model with Differentiable Duration Modeling

Keon Lee 170 Dec 27, 2022
Evaluation framework for testing segmentation networks in PyTorch

Evaluation framework for testing segmentation networks in PyTorch. What segmentation network to choose for next Kaggle competition? This benchmark knows the answer!

Eugene Khvedchenya 37 Apr 27, 2022
QQ Browser 2021 AI Algorithm Competition Track 1 1st Place Program

QQ Browser 2021 AI Algorithm Competition Track 1 1st Place Program

249 Jan 03, 2023