HAConvDR: History-Aware Conversational Dense Retrieval
Overview
- Paper: History-Aware Conversational Dense Retrieval
- Authors: Fengran Mo, Chen Qu, Kelong Mao, Tianyu Zhu, Zhan Su, Kaiyu Huang, et al. (7 authors)
- Published: 2024-01-30
- Source: https://arxiv.org/abs/2401.16659
- Category: Conversational Search
- A Survey of Conversational Search (ACM, Sep 2025)
- Engineering Conversational Search Systems: A Review (arXiv:2407.00997)
- Agentic Conversational Search with Contextualized Reasoning via Reinforcement Learning
- CTR-Guided Generative Query Suggestion in Conversational Search (EMNLP)
- ChatRetriever: Adapting Large Language Models for Generalized and Robust Conversational Dense Retrieval (arXiv:2404.13556)
- CoSearchAgent: A Lightweight Collaborative Search Agent (arXiv:2402.06363)
- Original paper: History-Aware Conversational Dense Retrieval. arXiv:2401.16659
Problem Statement
Conversational search facilitates complex information retrieval by enabling multi-turn interactions between users and the system. Supporting such interactions requires a comprehensive understanding of the conversational inputs to formulate a good search query based on historical information. In particular, the search query should include the relevant information from the previous conversation turns.
However, current approaches for conversational dense retrieval primarily rely on fine-tuning a pre-trained ad-hoc retriever using the whole conversational search session, which can be lengthy and noisy. Moreover, existing approaches are limited by the amount of manual supervision signals in the existing datasets.
Proposed Approach
The authors propose HAConvDR, a History-Aware Conversational Dense Retrieval system built on two ideas:
1. Context-denoised query reformulation — reducing the noise introduced by lengthy conversation histories when reformulating the current query. 2. Automatic mining of supervision signals — generating training signals automatically based on the actual impact of historical turns, instead of relying solely on limited manual annotations.
Results
Experiments on two public conversational search datasets demonstrate the improved history modeling capability of HAConvDR, particularly for long conversations with topic shifts.
Original Abstract
> Conversational search facilitates complex information retrieval by enabling multi-turn interactions between users and the system. Supporting such interactions requires a comprehensive understanding of the conversational inputs to formulate a good search query based on historical information. In particular, the search query should include the relevant information from the previous conversation turns. However, current approaches for conversational dense retrieval primarily rely on fine-tuning a pre-trained ad-hoc retriever using the whole conversational search session, which can be lengthy and noisy. Moreover, existing approaches are limited by the amount of manual supervision signals in the existing datasets. To address the aforementioned issues, we propose a History-Aware Conversational Dense Retrieval (HAConvDR) system, which incorporates two ideas: context-denoised query reformulation and automatic mining of supervision signals based on the actual impact of historical turns. Experiments on two public conversational search datasets demonstrate the improved history modeling capability of HAConvDR, in particular for long conversations with topic shifts.