Research

Paper

TESTING March 20, 2026

SaFRO: Satisfaction-Aware Fusion via Dual-Relative Policy Optimization for Short-Video Search

Authors

Renzhe Zhou, Songyang Li, Feiran Zhu, Chenglei Dai, Yi Zhang, Yi Wang, Jingwei Zhuo

Abstract

Multi-Task Fusion plays a pivotal role in industrial short-video search systems by aggregating heterogeneous prediction signals into a unified ranking score. However, existing approaches predominantly optimize for immediate engagement metrics, which often fail to align with long-term user satisfaction. While Reinforcement Learning (RL) offers a promising avenue for user satisfaction optimization, its direct application to search scenarios is non-trivial due to the inherent data sparsity and intent constraints compared to recommendation feeds. To this end, we propose SaFRO, a novel framework designed to optimize user satisfaction in short-video search. We first construct a satisfaction-aware reward model that utilizes query-level behavioral proxies to capture holistic user satisfaction beyond item-level interactions. Then we introduce Dual-Relative Policy Optimization (DRPO), an efficient policy learning method that updates the fusion policy through relative preference comparisons within groups and across batches. Furthermore, we design a Task-Relation-Aware Fusion module to explicitly model the interdependencies among different objectives, enabling context-sensitive weight adaptation. Extensive offline evaluations and large-scale online A/B tests on Kuaishou short-video search platform demonstrate that SaFRO significantly outperforms state-of-the-art baselines, delivering substantial gains in both short-term ranking quality and long-term user retention.

Metadata

arXiv ID: 2603.19585
Provider: ARXIV
Primary Category: cs.IR
Published: 2026-03-20
Fetched: 2026-03-23 16:54

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.19585v1</id>\n    <title>SaFRO: Satisfaction-Aware Fusion via Dual-Relative Policy Optimization for Short-Video Search</title>\n    <updated>2026-03-20T02:57:50Z</updated>\n    <link href='https://arxiv.org/abs/2603.19585v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.19585v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Multi-Task Fusion plays a pivotal role in industrial short-video search systems by aggregating heterogeneous prediction signals into a unified ranking score. However, existing approaches predominantly optimize for immediate engagement metrics, which often fail to align with long-term user satisfaction. While Reinforcement Learning (RL) offers a promising avenue for user satisfaction optimization, its direct application to search scenarios is non-trivial due to the inherent data sparsity and intent constraints compared to recommendation feeds. To this end, we propose SaFRO, a novel framework designed to optimize user satisfaction in short-video search. We first construct a satisfaction-aware reward model that utilizes query-level behavioral proxies to capture holistic user satisfaction beyond item-level interactions. Then we introduce Dual-Relative Policy Optimization (DRPO), an efficient policy learning method that updates the fusion policy through relative preference comparisons within groups and across batches. Furthermore, we design a Task-Relation-Aware Fusion module to explicitly model the interdependencies among different objectives, enabling context-sensitive weight adaptation. Extensive offline evaluations and large-scale online A/B tests on Kuaishou short-video search platform demonstrate that SaFRO significantly outperforms state-of-the-art baselines, delivering substantial gains in both short-term ranking quality and long-term user retention.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.IR'/>\n    <published>2026-03-20T02:57:50Z</published>\n    <arxiv:comment>9 pages, 8 figures</arxiv:comment>\n    <arxiv:primary_category term='cs.IR'/>\n    <author>\n      <name>Renzhe Zhou</name>\n    </author>\n    <author>\n      <name>Songyang Li</name>\n    </author>\n    <author>\n      <name>Feiran Zhu</name>\n    </author>\n    <author>\n      <name>Chenglei Dai</name>\n    </author>\n    <author>\n      <name>Yi Zhang</name>\n    </author>\n    <author>\n      <name>Yi Wang</name>\n    </author>\n    <author>\n      <name>Jingwei Zhuo</name>\n    </author>\n  </entry>"
}