![[2602.17667] When & How to Write for Personalized Demand-aware Query Rewriting in Video Search](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.17667] When & How to Write for Personalized Demand-aware Query Rewriting in Video Search
Summary
The paper presents WeWrite, a framework for personalized demand-aware query rewriting in video search, addressing challenges in user intent recognition and feedback delays.
Why It Matters
As video content continues to proliferate, enhancing search systems to accurately reflect user intent is crucial. WeWrite aims to improve user engagement by refining how queries are rewritten based on historical behavior, potentially transforming video search experiences.
Key Takeaways
- WeWrite identifies when personalization is necessary through automated mining of user logs.
- It employs a hybrid training approach combining Supervised Fine-Tuning and Group Relative Policy Optimization.
- The framework's architecture ensures low latency during deployment.
- A/B testing shows WeWrite improves video engagement metrics.
- The research highlights the importance of user behavior in enhancing search systems.
Computer Science > Information Retrieval arXiv:2602.17667 (cs) [Submitted on 17 Dec 2025] Title:When & How to Write for Personalized Demand-aware Query Rewriting in Video Search Authors:Cheng cheng, Chenxing Wang, Aolin Li, Haijun Wu, Huiyun Hu, Juyuan Wang View a PDF of the paper titled When & How to Write for Personalized Demand-aware Query Rewriting in Video Search, by Cheng cheng and 4 other authors View PDF HTML (experimental) Abstract:In video search systems, user historical behaviors provide rich context for identifying search intent and resolving ambiguity. However, traditional methods utilizing implicit history features often suffer from signal dilution and delayed feedback. To address these challenges, we propose WeWrite, a novel Personalized Demand-aware Query Rewriting framework. Specifically, WeWrite tackles three key challenges: (1) When to Write: An automated posterior-based mining strategy extracts high-quality samples from user logs, identifying scenarios where personalization is strictly necessary; (2) How to Write: A hybrid training paradigm combines Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO) to align the LLM's output style with the retrieval system; (3) Deployment: A parallel "Fake Recall" architecture ensures low latency. Online A/B testing on a large-scale video platform demonstrates that WeWrite improves the Click-Through Video Volume (VV$>$10s) by 1.07% and reduces the Query Reformulation Rate by 2.97%. Subjects: Inf...
