Automatic In-Domain Exemplar Construction and LLM-Based Refinement of Multi-LLM Expansions for Query Expansion
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866908884497596416 |
|---|---|
| author | Li, Minghan Nie, Ercong Zhao, Siqi Chen, Tongna Huang, Huiping Zhou, Guodong |
| author_facet | Li, Minghan Nie, Ercong Zhao, Siqi Chen, Tongna Huang, Huiping Zhou, Guodong |
| contents | Query expansion with large language models is promising but often relies on hand-crafted prompts, manually chosen exemplars, or a single LLM, making it non-scalable and sensitive to domain shift. We present an automated, domain-adaptive QE framework that builds in-domain exemplar pools by harvesting pseudo-relevant passages using a BM25-MonoT5 pipeline. A training-free cluster-based strategy selects diverse demonstrations, yielding strong and stable in-context QE without supervision. To further exploit model complementarity, we introduce a two-LLM ensemble in which two heterogeneous LLMs independently generate expansions and a refinement LLM consolidates them into one coherent expansion. Across TREC DL20, DBPedia, and SciFact, the refined ensemble delivers consistent and statistically significant gains over BM25, Rocchio, zero-shot, and fixed few-shot baselines. The framework offers a reproducible testbed for exemplar selection and multi-LLM generation, and a practical, label-free solution for real-world QE. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_08917 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Automatic In-Domain Exemplar Construction and LLM-Based Refinement of Multi-LLM Expansions for Query Expansion Li, Minghan Nie, Ercong Zhao, Siqi Chen, Tongna Huang, Huiping Zhou, Guodong Information Retrieval Artificial Intelligence Query expansion with large language models is promising but often relies on hand-crafted prompts, manually chosen exemplars, or a single LLM, making it non-scalable and sensitive to domain shift. We present an automated, domain-adaptive QE framework that builds in-domain exemplar pools by harvesting pseudo-relevant passages using a BM25-MonoT5 pipeline. A training-free cluster-based strategy selects diverse demonstrations, yielding strong and stable in-context QE without supervision. To further exploit model complementarity, we introduce a two-LLM ensemble in which two heterogeneous LLMs independently generate expansions and a refinement LLM consolidates them into one coherent expansion. Across TREC DL20, DBPedia, and SciFact, the refined ensemble delivers consistent and statistically significant gains over BM25, Rocchio, zero-shot, and fixed few-shot baselines. The framework offers a reproducible testbed for exemplar selection and multi-LLM generation, and a practical, label-free solution for real-world QE. |
| title | Automatic In-Domain Exemplar Construction and LLM-Based Refinement of Multi-LLM Expansions for Query Expansion |
| topic | Information Retrieval Artificial Intelligence |
| url | https://arxiv.org/abs/2602.08917 |