DS@GT at TREC TOT 2025: Bridging Vague Recollection with Fusion Retrieval and Learned Reranking
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914328698945536 |
|---|---|
| author | Zhou, Wenxin Mehta, Ritesh Miyaguchi, Anthony |
| author_facet | Zhou, Wenxin Mehta, Ritesh Miyaguchi, Anthony |
| contents | We develop a two-stage retrieval system that combines multiple complementary retrieval methods with a learned reranker and LLM-based reranking, to address the TREC Tip-of-the-Tongue (ToT) task. In the first stage, we employ hybrid retrieval that merges LLM-based retrieval, sparse (BM25), and dense (BGE-M3) retrieval methods. We also introduce topic-aware multi-index dense retrieval that partitions the Wikipedia corpus into 24 topical domains. In the second stage, we evaluate both a trained LambdaMART reranker and LLM-based reranking. To support model training, we generate 5000 synthetic ToT queries using LLMs. Our best system achieves recall of 0.66 and NDCG@1000 of 0.41 on the test set by combining hybrid retrieval with Gemini-2.5-flash reranking, demonstrating the effectiveness of fusion retrieval. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_15518 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DS@GT at TREC TOT 2025: Bridging Vague Recollection with Fusion Retrieval and Learned Reranking Zhou, Wenxin Mehta, Ritesh Miyaguchi, Anthony Information Retrieval Computation and Language Machine Learning We develop a two-stage retrieval system that combines multiple complementary retrieval methods with a learned reranker and LLM-based reranking, to address the TREC Tip-of-the-Tongue (ToT) task. In the first stage, we employ hybrid retrieval that merges LLM-based retrieval, sparse (BM25), and dense (BGE-M3) retrieval methods. We also introduce topic-aware multi-index dense retrieval that partitions the Wikipedia corpus into 24 topical domains. In the second stage, we evaluate both a trained LambdaMART reranker and LLM-based reranking. To support model training, we generate 5000 synthetic ToT queries using LLMs. Our best system achieves recall of 0.66 and NDCG@1000 of 0.41 on the test set by combining hybrid retrieval with Gemini-2.5-flash reranking, demonstrating the effectiveness of fusion retrieval. |
| title | DS@GT at TREC TOT 2025: Bridging Vague Recollection with Fusion Retrieval and Learned Reranking |
| topic | Information Retrieval Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2601.15518 |