Comprehensive Comparison of RAG Methods Across Multi-Domain Conversational QA

TL;DR

This paper systematically compares various retrieval-augmented generation methods for multi-turn conversational question answering across diverse datasets, revealing that simpler, well-aligned retrieval strategies often outperform complex techniques.

Contribution

It provides the first comprehensive empirical evaluation of RAG methods in multi-turn conversational QA, highlighting the importance of retrieval strategy alignment over complexity.

Findings

Reranking, hybrid BM25, and HyDE outperform vanilla RAG.

Advanced techniques may degrade performance below baseline.

Dataset characteristics and dialogue length significantly affect retrieval effectiveness.

Abstract

Conversational question answering increasingly relies on retrieval-augmented generation (RAG) to ground large language models (LLMs) in external knowledge. Yet, most existing studies evaluate RAG methods in isolation and primarily focus on single-turn settings. This paper addresses the lack of a systematic comparison of RAG methods for multi-turn conversational QA, where dialogue history, coreference, and shifting user intent substantially complicate retrieval. We present a comprehensive empirical study of vanilla and advanced RAG methods across eight diverse conversational QA datasets spanning multiple domains. Using a unified experimental setup, we evaluate retrieval quality and answer generation using generator and retrieval metrics, and analyze how performance evolves across conversation turns. Our results show that robust yet straightforward methods, such as reranking, hybrid BM25, and HyDE, consistently outperform vanilla RAG. In contrast, several advanced techniques fail to yield gains and can even degrade performance below the No-RAG baseline. We further demonstrate that dataset characteristics and dialogue length strongly influence retrieval effectiveness, explaining why no single RAG strategy dominates across settings. Overall, our findings indicate that effective conversational RAG depends less on method complexity than on alignment between the retrieval strategy and the dataset structure. We publish the code used.\footnote{\href{https://github.com/Klejda-A/exp-rag.git}{GitHub Repository}}