Passage Segmentation of Documents for Extractive Question Answering
This addresses the chunking bottleneck in RAG pipelines for question answering, but it is incremental as it focuses on optimizing an existing component.
The study tackled the problem of insufficient attention to chunking in Retrieval-Augmented Generation (RAG) for open-domain question answering, and the result was that the proposed Logits-Guided Multi-Granular Chunker (LGMGC) improved retrieval and outperformed existing chunking methods on two benchmark datasets.
Retrieval-Augmented Generation (RAG) has proven effective in open-domain question answering. However, the chunking process, which is essential to this pipeline, often receives insufficient attention relative to retrieval and synthesis components. This study emphasizes the critical role of chunking in improving the performance of both dense passage retrieval and the end-to-end RAG pipeline. We then introduce the Logits-Guided Multi-Granular Chunker (LGMGC), a novel framework that splits long documents into contextualized, self-contained chunks of varied granularity. Our experimental results, evaluated on two benchmark datasets, demonstrate that LGMGC not only improves the retrieval step but also outperforms existing chunking methods when integrated into a RAG pipeline.