Domain-specific Question Answering with Hybrid Search
Abstract
Domain specific question answering is an evolving field that requires specialized solutions to address unique challenges. In this paper, we show that a hybrid approach combining a fine-tuned dense retriever with keyword based sparse search methods significantly enhances performance. Our system leverages a linear combination of relevance signals, including cosine similarity from dense retrieval, BM25 scores, and URL host matching, each with tunable boost parameters. Experimental results indicate that this hybrid method outperforms our single-retriever system, achieving improved accuracy while maintaining robust contextual grounding. These findings suggest that integrating multiple retrieval methodologies with weighted scoring effectively addresses the complexities of domain specific question answering in enterprise settings.
Cite
@article{arxiv.2412.03736,
title = {Domain-specific Question Answering with Hybrid Search},
author = {Dewang Sultania and Zhaoyu Lu and Twisha Naik and Franck Dernoncourt and David Seunghyun Yoon and Sanat Sharma and Trung Bui and Ashok Gupta and Tushar Vatsa and Suhas Suresha and Ishita Verma and Vibha Belavadi and Cheng Chen and Michael Friedrich},
journal= {arXiv preprint arXiv:2412.03736},
year = {2024}
}
Comments
AAAI-25 Workshop on Document Understanding and Intelligence