Mitigating Matching Biases Through Score Calibration

• URL: http://arxiv.org/abs/2411.01685v1
• Date: Sun, 03 Nov 2024 21:01:40 GMT
• Title: Mitigating Matching Biases Through Score Calibration
• Authors: Mohammad Hossein Moslemi, Mostafa Milani,
• Abstract summary: Biased outcomes in record matching can result in unequal error rates across demographic groups, raising ethical and legal concerns. In this paper, we adapt fairness metrics traditionally applied in regression models to evaluate cumulative bias across all thresholds in record matching. We propose a novel post-processing calibration method, leveraging optimal transport theory and Wasserstein barycenters, to balance matching scores across demographic groups.
• Score: 1.5530839016602822
• License:
• Abstract: Record matching, the task of identifying records that correspond to the same real-world entities across databases, is critical for data integration in domains like healthcare, finance, and e-commerce. While traditional record matching models focus on optimizing accuracy, fairness issues, such as demographic disparities in model performance, have attracted increasing attention. Biased outcomes in record matching can result in unequal error rates across demographic groups, raising ethical and legal concerns. Existing research primarily addresses fairness at specific decision thresholds, using bias metrics like Demographic Parity (DP), Equal Opportunity (EO), and Equalized Odds (EOD) differences. However, threshold-specific metrics may overlook cumulative biases across varying thresholds. In this paper, we adapt fairness metrics traditionally applied in regression models to evaluate cumulative bias across all thresholds in record matching. We propose a novel post-processing calibration method, leveraging optimal transport theory and Wasserstein barycenters, to balance matching scores across demographic groups. This approach treats any matching model as a black box, making it applicable to a wide range of models without access to their training data. Our experiments demonstrate the effectiveness of the calibration method in reducing demographic parity difference in matching scores. To address limitations in reducing EOD and EO differences, we introduce a conditional calibration method, which empirically achieves fairness across widely used benchmarks and state-of-the-art matching methods. This work provides a comprehensive framework for fairness-aware record matching, setting the foundation for more equitable data integration processes.

Related papers

• MITA: Bridging the Gap between Model and Data for Test-time Adaptation [68.62509948690698]
  Test-Time Adaptation (TTA) has emerged as a promising paradigm for enhancing the generalizability of models. We propose Meet-In-The-Middle based MITA, which introduces energy-based optimization to encourage mutual adaptation of the model and data from opposing directions.
  arXiv  Detail & Related papers  (2024-10-12T07:02:33Z)
• Fair-OBNC: Correcting Label Noise for Fairer Datasets [9.427445881721814]
  biases in the training data are sometimes related to label noise. Models trained on such biased data may perpetuate or even aggravate the biases with respect to sensitive information. We propose Fair-OBNC, a label noise correction method with fairness considerations.
  arXiv  Detail & Related papers  (2024-10-08T17:18:18Z)
• Editable Fairness: Fine-Grained Bias Mitigation in Language Models [52.66450426729818]
  We propose a novel debiasing approach, Fairness Stamp (FAST), which enables fine-grained calibration of individual social biases. FAST surpasses state-of-the-art baselines with superior debiasing performance. This highlights the potential of fine-grained debiasing strategies to achieve fairness in large language models.
  arXiv  Detail & Related papers  (2024-08-07T17:14:58Z)
• Score Normalization for Demographic Fairness in Face Recognition [16.421833444307232]
  Well-known sample-centered score normalization techniques, Z-norm and T-norm, do not improve fairness for high-security operating points. We extend the standard Z/T-norm to integrate demographic information in normalization. We show that our techniques generally improve the overall fairness of five state-of-the-art pre-trained face recognition networks.
  arXiv  Detail & Related papers  (2024-07-19T07:51:51Z)
• Threshold-Independent Fair Matching through Score Calibration [1.5530839016602822]
  We introduce a new approach in Entity Matching (EM) using recent metrics for evaluating biases in score based binary classification. This approach enables the application of various bias metrics like equalized odds, equal opportunity, and demographic parity without depending on threshold settings. This paper contributes to the field of fairness in data cleaning, especially within EM, by promoting a method for generating matching scores that reduce biases across different thresholds.
  arXiv  Detail & Related papers  (2024-05-30T13:37:53Z)
• Systematic analysis of the impact of label noise correction on ML Fairness [0.0]
  We develop an empirical methodology to evaluate the effectiveness of label noise correction techniques in ensuring the fairness of models trained on biased datasets. Our results suggest that the Hybrid Label Noise Correction method achieves the best trade-off between predictive performance and fairness.
  arXiv  Detail & Related papers  (2023-06-28T08:08:14Z)
• Correcting Underrepresentation and Intersectional Bias for Classification [49.1574468325115]
  We consider the problem of learning from data corrupted by underrepresentation bias. We show that with a small amount of unbiased data, we can efficiently estimate the group-wise drop-out rates. We show that our algorithm permits efficient learning for model classes of finite VC dimension.
  arXiv  Detail & Related papers  (2023-06-19T18:25:44Z)
• Improving Fair Training under Correlation Shifts [33.385118640843416]
  In particular, when the bias between labels and sensitive groups changes, the fairness of the trained model is directly influenced and can worsen. We analytically show that existing in-processing fair algorithms have fundamental limits in accuracy and group fairness. We propose a novel pre-processing step that samples the input data to reduce correlation shifts.
  arXiv  Detail & Related papers  (2023-02-05T07:23:35Z)
• D-BIAS: A Causality-Based Human-in-the-Loop System for Tackling Algorithmic Bias [57.87117733071416]
  We propose D-BIAS, a visual interactive tool that embodies human-in-the-loop AI approach for auditing and mitigating social biases. A user can detect the presence of bias against a group by identifying unfair causal relationships in the causal network. For each interaction, say weakening/deleting a biased causal edge, the system uses a novel method to simulate a new (debiased) dataset.
  arXiv  Detail & Related papers  (2022-08-10T03:41:48Z)
• Balancing out Bias: Achieving Fairness Through Training Reweighting [58.201275105195485]
  Bias in natural language processing arises from models learning characteristics of the author such as gender and race. Existing methods for mitigating and measuring bias do not directly account for correlations between author demographics and linguistic variables. This paper introduces a very simple but highly effective method for countering bias using instance reweighting.
  arXiv  Detail & Related papers  (2021-09-16T23:40:28Z)
• Towards Model-Agnostic Post-Hoc Adjustment for Balancing Ranking Fairness and Algorithm Utility [54.179859639868646]
  Bipartite ranking aims to learn a scoring function that ranks positive individuals higher than negative ones from labeled data. There have been rising concerns on whether the learned scoring function can cause systematic disparity across different protected groups. We propose a model post-processing framework for balancing them in the bipartite ranking scenario.
  arXiv  Detail & Related papers  (2020-06-15T10:08:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.