SRAG: Structured Retrieval-Augmented Generation for Multi-Entity Question Answering over Wikipedia Graph
This addresses a specific bottleneck in retrieval-augmented generation for multi-entity QA, offering a domain-specific solution for Wikipedia-based tasks.
The paper tackles the problem of multi-entity question answering over Wikipedia graphs, where LLMs struggle to consolidate scattered information across documents, by introducing a structured RAG framework that organizes entities into relational tables and applies table-based reasoning. The result is a 29.6% improvement in accuracy over state-of-the-art methods.
Multi-entity question answering (MEQA) poses significant challenges for large language models (LLMs), which often struggle to consolidate scattered information across multiple documents. An example question might be "What is the distribution of IEEE Fellows among various fields of study?", which requires retrieving information from diverse sources e.g., Wikipedia pages. The effectiveness of current retrieval-augmented generation (RAG) methods is limited by the LLMs' capacity to aggregate insights from numerous pages. To address this gap, this paper introduces a structured RAG (SRAG) framework that systematically organizes extracted entities into relational tables (e.g., tabulating entities with schema columns like "name" and "field of study") and then apply table-based reasoning techniques. Our approach decouples retrieval and reasoning, enabling LLMs to focus on structured data analysis rather than raw text aggregation. Extensive experiments on Wikipedia-based multi-entity QA tasks demonstrate that SRAG significantly outperforms state-of-the-art long-context LLMs and RAG solutions, achieving a 29.6% improvement in accuracy. The results underscore the efficacy of structuring unstructured data to enhance LLMs' reasoning capabilities.