Large Language Model as Universal Retriever in Industrial-Scale Recommender System
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910952654372864 |
|---|---|
| author | Jiang, Junguang Huang, Yanwen Liu, Bin Kong, Xiaoyu Li, Xinhang Xu, Ziru Zhu, Han Xu, Jian Zheng, Bo |
| author_facet | Jiang, Junguang Huang, Yanwen Liu, Bin Kong, Xiaoyu Li, Xinhang Xu, Ziru Zhu, Han Xu, Jian Zheng, Bo |
| contents | In real-world recommender systems, different retrieval objectives are typically addressed using task-specific datasets with carefully designed model architectures. We demonstrate that Large Language Models (LLMs) can function as universal retrievers, capable of handling multiple objectives within a generative retrieval framework. To model complex user-item relationships within generative retrieval, we propose multi-query representation. To address the challenge of extremely large candidate sets in industrial recommender systems, we introduce matrix decomposition to boost model learnability, discriminability, and transferability, and we incorporate probabilistic sampling to reduce computation costs. Finally, our Universal Retrieval Model (URM) can adaptively generate a set from tens of millions of candidates based on arbitrary given objective while keeping the latency within tens of milliseconds. Applied to industrial-scale data, URM outperforms expert models elaborately designed for different retrieval objectives on offline experiments and significantly improves the core metric of online advertising platform by $3\%$. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_03041 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Large Language Model as Universal Retriever in Industrial-Scale Recommender System Jiang, Junguang Huang, Yanwen Liu, Bin Kong, Xiaoyu Li, Xinhang Xu, Ziru Zhu, Han Xu, Jian Zheng, Bo Information Retrieval Machine Learning In real-world recommender systems, different retrieval objectives are typically addressed using task-specific datasets with carefully designed model architectures. We demonstrate that Large Language Models (LLMs) can function as universal retrievers, capable of handling multiple objectives within a generative retrieval framework. To model complex user-item relationships within generative retrieval, we propose multi-query representation. To address the challenge of extremely large candidate sets in industrial recommender systems, we introduce matrix decomposition to boost model learnability, discriminability, and transferability, and we incorporate probabilistic sampling to reduce computation costs. Finally, our Universal Retrieval Model (URM) can adaptively generate a set from tens of millions of candidates based on arbitrary given objective while keeping the latency within tens of milliseconds. Applied to industrial-scale data, URM outperforms expert models elaborately designed for different retrieval objectives on offline experiments and significantly improves the core metric of online advertising platform by $3\%$. |
| title | Large Language Model as Universal Retriever in Industrial-Scale Recommender System |
| topic | Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2502.03041 |