Fair Ranking with Noisy Protected Attributes

Authors: Anay Mehrotra, Nisheeth Vishnoi

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Empirically, we observe that, compared to baselines, our algorithm outputs rankings with higher fairness, and has a similar or better fairness-utility trade-off compared to baselines. ... In this section we evaluate our framework s performance on synthetic and real-world data.
Researcher Affiliation Academia Anay Mehrotra Yale University Nisheeth K. Vishnoi Yale University
Pseudocode Yes Our algorithm (Algorithm 1) solves the standard linear programming relaxation of Program (7) to find a solution Rc and then uses a dependent-rounding algorithm by [19] to convert Rc to a ranking.
Open Source Code Yes Code for our simulations is available at https://github.com/AnayMehrotra/FairRankingWithNoisyAttributes
Open Datasets Yes We use the Occupations dataset [15]... We consider the chess ranking data [29]
Dataset Splits No The paper mentions parameters like m=500, n=25, and varies phi, but does not explicitly provide information on train/validation/test splits by percentage, sample counts, or reference predefined splits for reproducibility. It only states "We perform this calibration once and on all occupations and, then, use it for gender-stereotypical occupations" for data usage.
Hardware Specification No The paper does not provide specific hardware details (e.g., GPU/CPU models, memory, cloud instance types) used for running the experiments.
Software Dependencies No The paper mentions using a "CNN-based gender-classifier f [59]" but does not specify the version of this classifier or any other software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow versions).
Experiment Setup Yes Setup. We consider the DCG model of utilities [33] and a relaxation of equal representation constraints: (1) Given an intrinsic value wi 0, for each item i, we set Wij := wi (log (j + 1)) 1 j [n]. (2) Given a parameter ϕ [1, p], we set upper bounds Ukℓ:= ϕ p k for each k [n] and ℓ [p]. In simulations, we set m = 500, n = 25, and vary ϕ from p to 1. ... As a heuristic, we set γk = 1 20 maxℓ [p] q 1 Ukℓ in all simulations.