KERAG: Knowledge-Enhanced Retrieval-Augmented Generation for Advanced Question Answering

2025-09-04

AI Summary

Traditional knowledge graph question answering (KGQA) methods suffer from rigid schema constraints and semantic ambiguity, resulting in limited knowledge coverage and weak multi-hop reasoning capabilities. To address these limitations, we propose a knowledge graph-enhanced retrieval-augmented generation (KG-RAG) framework. Our approach replaces brittle semantic parsing with a flexible subgraph retrieval mechanism, integrated with subgraph filtering, information summarization, and large language model (LLM)-fine-tuned chain-of-thought reasoning to suppress noise and improve accuracy. This end-to-end pipeline balances precision and generalizability, outperforming state-of-the-art methods by approximately 7% across multiple benchmarks and achieving 10–21% higher answer quality than GPT-4o (Tool). The core innovation lies in decoupling knowledge retrieval from schema dependency—thereby substantially expanding coverage and strengthening complex, multi-hop reasoning.

Technology Category

• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Knowledge Representation and Reasoning: Knowledge Acquisition
• Natural Language Processing: Question Answering

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Retrieval-Augmented Generation (RAG) mitigates hallucination in Large Language Models (LLMs) by incorporating external data, with Knowledge Graphs (KGs) offering crucial information for question answering. Traditional Knowledge Graph Question Answering (KGQA) methods rely on semantic parsing, which typically retrieves knowledge strictly necessary for answer generation, thus often suffer from low coverage due to rigid schema requirements and semantic ambiguity. We present KERAG, a novel KG-based RAG pipeline that enhances QA coverage by retrieving a broader subgraph likely to contain relevant information. Our retrieval-filtering-summarization approach, combined with fine-tuned LLMs for Chain-of-Thought reasoning on knowledge sub-graphs, reduces noises and improves QA for both simple and complex questions. Experiments demonstrate that KERAG surpasses state-of-the-art solutions by about 7% in quality and exceeds GPT-4o (Tool) by 10-21%.

Problem

• Enhancing question answering coverage using knowledge graphs
• Reducing semantic ambiguity in retrieval-augmented generation systems
• Improving complex question handling through subgraph retrieval

Innovation

• Retrieves broader knowledge subgraphs for coverage
• Uses retrieval-filtering-summarization with fine-tuned LLMs
• Applies Chain-of-Thought reasoning on knowledge sub-graphs