Reinforcement Learning to Rank Using Coarse-grained Rewards
ACM Transactions on Information Systems · 12.11.2026
DOI: 10.1145/3862850 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Information Systems, online erschienen am 12.11.2026. Verfasst von Yiteng Tu und 6 weiteren Autor:innen. Die Arbeit nennt 87 Literaturangaben.
Abstract (Originalsprache)
Learning to rank (LTR) plays a crucial role in various Information Retrieval (IR) tasks. Although supervised LTR methods based on fine-grained relevance labels (e.g., document-level annotations) have achieved significant success, their reliance on costly and potentially biased annotations limits scalability and alignment with realistic goals. In contrast, coarse-grained feedback signals, such as duration time and session-level engagement, are more accessible and affordable. Reinforcement Learning (RL) offers a promising framework to directly optimize these objectives using reward signals, but most existing Reinforcement Learning to Rank (RLTR) approaches suffer from high variance and low sample efficiency. Motivated by recent advances in large language models (LLMs), we re-examine the problem of RLTR with coarse-grained rewards and propose new RLTR methods based on widely used RL algorithms for LLMs. We systematically compare supervised learning and RL-based methods across various model architectures and coarse-grained reward functions on large-scale LTR benchmarks. Experimental results demonstrate that advanced RL methods can directly learn from coarse-grained rewards and outperform strong supervised learning baselines even with fine-grained labels. This shows the great potential of RLTR for metric-agnostic ranking optimization. We release our code at: https://github.com/StibiumT16/RLTR .
Zitieren
APA
Tu, Y., Xu, Z., Yang, T., Su, W., Zhou, Y., Liu, Y., & Ai, Q. (2026). Reinforcement Learning to Rank Using Coarse-grained Rewards. ACM Transactions on Information Systems. https://doi.org/10.1145/3862850
Vancouver
Tu Y, Xu Z, Yang T, Su W, Zhou Y, Liu Y, et al. Reinforcement Learning to Rank Using Coarse-grained Rewards. ACM Transactions on Information Systems. 2026. doi:10.1145/3862850
BibTeX
@article{Tu2026,
author = {Tu, Yiteng and Xu, Zhichao and Yang, Tao and Su, Weihang and Zhou, Yujia and Liu, Yiqun and Ai, Qingyao},
title = {Reinforcement Learning to Rank Using Coarse-grained Rewards},
journal = {ACM Transactions on Information Systems},
year = {2026},
doi = {10.1145/3862850}
}
EUVIMED