Continuous Fairness On Data Streams

2026-01-13

AI Summary

This work addresses the challenge of maintaining fine-grained group fairness in sliding-window data streams, where existing methods struggle to provide consistent guarantees. The authors propose the first block-level group fairness model, which partitions the sliding window into finer-grained blocks to enforce localized fairness constraints. To efficiently track attribute distributions, they design a sketch-based data structure that enables real-time monitoring with low overhead. Furthermore, they introduce a theoretically optimal online reordering algorithm that rapidly corrects output rankings upon detecting fairness violations. Experimental results demonstrate that the system achieves millisecond-level average latency and a throughput of 30,000 queries per second, while improving block-level group fairness by 50–60% on average, with gains reaching up to 95% in the best case.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Data Mining & Knowledge Management: Data Stream Mining
• Search and Optimization: Distributed Search

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Fairness and ethical considerations in crowd work and in human-in-the-loop AI systems

Abstract

We study the problem of enforcing continuous group fairness over windows in data streams. We propose a novel fairness model that ensures group fairness at a finer granularity level (referred to as block) within each sliding window. This formulation is particularly useful when the window size is large, making it desirable to enforce fairness at a finer granularity. Within this framework, we address two key challenges: efficiently monitoring whether each sliding window satisfies block-level group fairness, and reordering the current window as effectively as possible when fairness is violated. To enable real-time monitoring, we design sketch-based data structures that maintain attribute distributions with minimal overhead. We also develop optimal, efficient algorithms for the reordering task, supported by rigorous theoretical guarantees. Our evaluation on four real-world streaming scenarios demonstrates the practical effectiveness of our approach. We achieve millisecond-level processing and a throughput of approximately 30,000 queries per second on average, depending on system parameters. The stream reordering algorithm improves block-level group fairness by up to 95% in certain cases, and by 50-60% on average across datasets. A qualitative study further highlights the advantages of block-level fairness compared to window-level fairness.

Problem

• continuous fairness
• data streams
• group fairness
• sliding window
• block-level fairness

Innovation

• continuous fairness
• data streams
• block-level fairness
• sketch-based monitoring
• stream reordering