CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation

Chao-Wei Huang, Chen-Yu Hsu, Tsu-Yuan Hsu, Chen-An Li, Yun-Nung Chen
9/13/2023
cs.CLcs.IR

Official code: MiuLab/CONVERSER (8 stars, last commit 2024-02-05).

Abstract

Conversational search provides a natural interface for information retrieval (IR). Recent approaches have demonstrated promising results in applying dense retrieval to conversational IR. However, training dense retrievers requires large amounts of in-domain paired data. This hinders the development of conversational dense retrievers, as abundant in-domain conversations are expensive to collect. In this paper, we propose CONVERSER, a framework for training conversational dense retrievers with at most 6 examples of in-domain dialogues. Specifically, we utilize the in-context learning capability of large language models to generate conversational queries given a passage in the retrieval corpus. Experimental results on conversational retrieval benchmarks OR-QuAC and TREC CAsT 19 show that the proposed CONVERSER achieves comparable performance to fully-supervised models, demonstrating the effectiveness of our proposed framework in few-shot conversational dense retrieval. All source code and generated datasets are available at https://github.com/MiuLab/CONVERSER

View on arXivView PDF

Code Implementations(1)

MiuLab/CONVERSEROfficial100%

CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation, SIGDIAL 2023

81PythonJul 28, 20232 years ago
conversational-aidense-retrievalsigdial2023synthetic-data

Similar Papers

DiSCo: LLM Knowledge Distillation for Efficient Sparse Retrieval in Conversational Search

87% match1 repo72024

SSP: Self-Supervised Post-training for Conversational Search

87% match2023

ConvSDG: Session Data Generation for Conversational Search

86% match1 repo42024

ConvMix: A Mixed-Criteria Data Augmentation Framework for Conversational Dense Retrieval

86% match1 repo2025

UniConv: Unifying Retrieval and Response Generation for Large Language Models in Conversations

86% match2025

Generalizing Conversational Dense Retrieval via LLM-Cognition Data Augmentation

85% match1 repo72024

Cite this paper

@article{huang2023converser,
  title  = {CONVERSER: Few-Shot Conversational Dense Retrieval with Synthetic Data Generation},
  author = {Chao-Wei Huang and Chen-Yu Hsu and Tsu-Yuan Hsu and Chen-An Li and Yun-Nung Chen},
  year   = {2023},
  eprint = {2309.06748},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2309.06748v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2309.06748.svg)](https://codeofpaper.com/paper/2309.06748)

Discussion