1-800-SHARED-TASKS at RegNLP: Lexical Reranking of Semantic Retrieval (LeSeR) for Regulatory Question Answering
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916513374535680 |
|---|---|
| author | Purbey, Jebish Sharma, Drishti Gupta, Siddhant Murad, Khawaja Pullakhandam, Siddartha Kadiyala, Ram Mohan Rao |
| author_facet | Purbey, Jebish Sharma, Drishti Gupta, Siddhant Murad, Khawaja Pullakhandam, Siddartha Kadiyala, Ram Mohan Rao |
| contents | This paper presents the system description of our entry for the COLING 2025 RegNLP RIRAG (Regulatory Information Retrieval and Answer Generation) challenge, focusing on leveraging advanced information retrieval and answer generation techniques in regulatory domains. We experimented with a combination of embedding models, including Stella, BGE, CDE, and Mpnet, and leveraged fine-tuning and reranking for retrieving relevant documents in top ranks. We utilized a novel approach, LeSeR, which achieved competitive results with a recall@10 of 0.8201 and map@10 of 0.6655 for retrievals. This work highlights the transformative potential of natural language processing techniques in regulatory applications, offering insights into their capabilities for implementing a retrieval augmented generation system while identifying areas for future improvement in robustness and domain adaptation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_06009 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | 1-800-SHARED-TASKS at RegNLP: Lexical Reranking of Semantic Retrieval (LeSeR) for Regulatory Question Answering Purbey, Jebish Sharma, Drishti Gupta, Siddhant Murad, Khawaja Pullakhandam, Siddartha Kadiyala, Ram Mohan Rao Computation and Language Information Retrieval Machine Learning This paper presents the system description of our entry for the COLING 2025 RegNLP RIRAG (Regulatory Information Retrieval and Answer Generation) challenge, focusing on leveraging advanced information retrieval and answer generation techniques in regulatory domains. We experimented with a combination of embedding models, including Stella, BGE, CDE, and Mpnet, and leveraged fine-tuning and reranking for retrieving relevant documents in top ranks. We utilized a novel approach, LeSeR, which achieved competitive results with a recall@10 of 0.8201 and map@10 of 0.6655 for retrievals. This work highlights the transformative potential of natural language processing techniques in regulatory applications, offering insights into their capabilities for implementing a retrieval augmented generation system while identifying areas for future improvement in robustness and domain adaptation. |
| title | 1-800-SHARED-TASKS at RegNLP: Lexical Reranking of Semantic Retrieval (LeSeR) for Regulatory Question Answering |
| topic | Computation and Language Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2412.06009 |