Bridging the Long-Tail Gap: Robust Retrieval-Augmented Relation Completion via Multi-Stage Paraphrase Infusion
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910162063720448 |
|---|---|
| author | Alam, Fahmida Surdeanu, Mihai Riloff, Ellen |
| author_facet | Alam, Fahmida Surdeanu, Mihai Riloff, Ellen |
| contents | Large language models (LLMs) struggle with relation completion (RC), both with and without retrieval-augmented generation (RAG), particularly when the required information is rare or sparsely represented. To address this, we propose a novel multi-stage paraphrase-guided relation-completion framework, RC-RAG, that systematically incorporates relation paraphrases across multiple stages. In particular, RC-RAG: (a) integrates paraphrases into retrieval to expand lexical coverage of the relation, (b) uses paraphrases to generate relation-aware summaries, and (c) leverages paraphrases during generation to guide reasoning for relation completion. Importantly, our method does not require any model fine-tuning. Experiments with five LLMs on two benchmark datasets show that RC-RAG consistently outperforms several RAG baselines. In long-tail settings, the best-performing LLM augmented with RC-RAG improves by 40.6 Exact Match (EM) points over its standalone performance and surpasses two strong RAG baselines by 16.0 and 13.8 EM points, respectively, while maintaining low computational overhead. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_22261 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Bridging the Long-Tail Gap: Robust Retrieval-Augmented Relation Completion via Multi-Stage Paraphrase Infusion Alam, Fahmida Surdeanu, Mihai Riloff, Ellen Computation and Language Large language models (LLMs) struggle with relation completion (RC), both with and without retrieval-augmented generation (RAG), particularly when the required information is rare or sparsely represented. To address this, we propose a novel multi-stage paraphrase-guided relation-completion framework, RC-RAG, that systematically incorporates relation paraphrases across multiple stages. In particular, RC-RAG: (a) integrates paraphrases into retrieval to expand lexical coverage of the relation, (b) uses paraphrases to generate relation-aware summaries, and (c) leverages paraphrases during generation to guide reasoning for relation completion. Importantly, our method does not require any model fine-tuning. Experiments with five LLMs on two benchmark datasets show that RC-RAG consistently outperforms several RAG baselines. In long-tail settings, the best-performing LLM augmented with RC-RAG improves by 40.6 Exact Match (EM) points over its standalone performance and surpasses two strong RAG baselines by 16.0 and 13.8 EM points, respectively, while maintaining low computational overhead. |
| title | Bridging the Long-Tail Gap: Robust Retrieval-Augmented Relation Completion via Multi-Stage Paraphrase Infusion |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2604.22261 |