Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.19585 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911530958716928 |
|---|---|
| author | Zhou, Renzhe Li, Songyang Zhu, Feiran Dai, Chenglei Zhang, Yi Wang, Yi Zhuo, Jingwei |
| author_facet | Zhou, Renzhe Li, Songyang Zhu, Feiran Dai, Chenglei Zhang, Yi Wang, Yi Zhuo, Jingwei |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_19585 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | SaFRO: Satisfaction-Aware Fusion via Dual-Relative Policy Optimization for Short-Video Search Zhou, Renzhe Li, Songyang Zhu, Feiran Dai, Chenglei Zhang, Yi Wang, Yi Zhuo, Jingwei Information Retrieval 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. |
| title | SaFRO: Satisfaction-Aware Fusion via Dual-Relative Policy Optimization for Short-Video Search |
| topic | Information Retrieval |
| url | https://arxiv.org/abs/2603.19585 |