GE-Chat: A Graph Enhanced RAG Framework for Evidential Response Generation of LLMs

2025-05-15

AI Summary

To address the low trustworthiness, hallucination propensity, and lack of verifiable evidence support in large language models (LLMs), this paper proposes a knowledge graph–enhanced retrieval-augmented generation (RAG) framework. The framework enables users to upload domain-specific documents and automatically constructs a corresponding knowledge graph. It introduces two key innovations: (1) a graph-structure-driven *n*-hop subgraph retrieval mechanism and (2) an entailment-aware sentence generation module, ensuring explicit alignment between generated answers and original evidential sources. By integrating chain-of-thought (CoT) reasoning with entailment modeling, the framework significantly improves LLMs’ ability to precisely locate, trace, and verify evidence within free-text corpora. Experimental results demonstrate superior performance over baseline models in response credibility, interpretability, and user trust. This work establishes a novel paradigm for building auditable, verifiable, and trustworthy LLM systems.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust
• Natural Language Processing: (Large) Language Models

Application Category

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

Abstract

Large Language Models are now key assistants in human decision-making processes. However, a common note always seems to follow:"LLMs can make mistakes. Be careful with important info."This points to the reality that not all outputs from LLMs are dependable, and users must evaluate them manually. The challenge deepens as hallucinated responses, often presented with seemingly plausible explanations, create complications and raise trust issues among users. To tackle such issue, this paper proposes GE-Chat, a knowledge Graph enhanced retrieval-augmented generation framework to provide Evidence-based response generation. Specifically, when the user uploads a material document, a knowledge graph will be created, which helps construct a retrieval-augmented agent, enhancing the agent's responses with additional knowledge beyond its training corpus. Then we leverage Chain-of-Thought (CoT) logic generation, n-hop sub-graph searching, and entailment-based sentence generation to realize accurate evidence retrieval. We demonstrate that our method improves the existing models' performance in terms of identifying the exact evidence in a free-form context, providing a reliable way to examine the resources of LLM's conclusion and help with the judgment of the trustworthiness.

Problem

• Enhancing LLM response reliability with evidence-based generation
• Reducing hallucinations in LLM outputs via knowledge graphs
• Improving trustworthiness of LLM conclusions through structured retrieval

Innovation

• Graph-enhanced RAG framework for evidence-based responses
• Knowledge graph creation from user-uploaded documents
• CoT logic, n-hop search, entailment-based generation