Fairness and Bias in Truth Discovery Algorithms: An Experimental Analysis

TL;DR

This paper systematically analyzes bias and fairness issues in truth discovery algorithms used for aggregating crowd-labeled data, revealing biases from workers and limitations of existing methods, and advocates for developing bias-aware solutions.

Contribution

It provides the first comprehensive empirical study of bias in truth discovery algorithms and highlights their limitations in ensuring fair and unbiased labels.

Findings

A significant proportion of workers provide biased labels.

Simple truth discovery methods are sub-optimal in mitigating bias.

Existing fairness correction methods are ineffective in this context.

Abstract

Machine learning (ML) based approaches are increasingly being used in a number of applications with societal impact. Training ML models often require vast amounts of labeled data, and crowdsourcing is a dominant paradigm for obtaining labels from multiple workers. Crowd workers may sometimes provide unreliable labels, and to address this, truth discovery (TD) algorithms such as majority voting are applied to determine the consensus labels from conflicting worker responses. However, it is important to note that these consensus labels may still be biased based on sensitive attributes such as gender, race, or political affiliation. Even when sensitive attributes are not involved, the labels can be biased due to different perspectives of subjective aspects such as toxicity. In this paper, we conduct a systematic study of the bias and fairness of TD algorithms. Our findings using two existing crowd-labeled datasets, reveal that a non-trivial proportion of workers provide biased results, and using simple approaches for TD is sub-optimal. Our study also demonstrates that popular TD algorithms are not a panacea. Additionally, we quantify the impact of these unfair workers on downstream ML tasks and show that conventional methods for achieving fairness and correcting label biases are ineffective in this setting. We end the paper with a plea for the design of novel bias-aware truth discovery algorithms that can ameliorate these issues.