Reproducibility Companion Paper:In-processing User Constrained Dominant Sets for User-Oriented Fairness in Recommender Systems

TL;DR

The paper tackles user-oriented fairness in recommender systems by proposing In-processing User Constrained Dominant Sets (In-UCDS), a two-stage framework that first performs constrained clustering of disadvantaged users with similar advantaged peers via Dominant Sets and then incorporates a fairness loss into the training objective to guide fair learning. By reproducing experiments on three real-world datasets using four backbone models and multiple baselines, the work demonstrates that In-UCDS can improve fairness (lower $M_{UOF}$) without sacrificing, and often improving, ranking performance (higher $NDCG$ and $F1$). The authors provide comprehensive artifacts—datasets, source code, configuration files, and environment details—to facilitate exact reproduction, thereby enhancing transparency and reliability in fairness research for recommender systems. This reproducibility companion offers practical guidance for researchers to replicate results and extend the In-UCDS approach in diverse domains and settings, contributing to more equitable user experiences in personalized recommendations.

Abstract

In this paper, we reproduce experimental results presented in our earlier work titled "In-processing User Constrained Dominant Sets for User-Oriented Fairness in Recommender Systems" that was presented in the proceeding of the 31st ACM International Conference on Multimedia.This work aims to verify the effectiveness of our previously proposed method and provide guidance for reproducibility. We present detailed descriptions of our preprocessed datasets, the structure of our source code, configuration file settings, experimental environment, and the reproduced experimental results.