RAGulating Compliance: A Multi-Agent Knowledge Graph for Regulatory QA

2025-08-13

AI Summary

Large language models (LLMs) exhibit low factual accuracy and weak verifiability in regulatory compliance question answering. Method: This paper proposes an ontology-free, dynamic multi-agent knowledge graph framework integrating subject-predicate-object (SPO) triplet extraction, knowledge graph embedding, and retrieval-augmented generation (RAG). Multiple specialized agents collaboratively perform structured parsing of regulatory documents, triplet cleaning, and incremental graph updates, while a unified vector database enables joint graph-structured reasoning and semantic retrieval. Contribution/Results: The framework introduces novel capabilities—including semantic alignment for complex queries, subgraph visualization, and audit-grade provenance tracing. Experiments demonstrate significant improvements over conventional RAG and fine-tuning baselines in factual accuracy, interpretability, and robustness. It establishes a new high-trust paradigm for compliance QA in highly regulated domains such as finance and healthcare.

Technology Category

• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Machine Learning: Large Multimodal Models (LMMs)
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

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

Regulatory compliance question answering (QA) requires precise, verifiable information, and domain-specific expertise, posing challenges for Large Language Models (LLMs). In this work, we present a novel multi-agent framework that integrates a Knowledge Graph (KG) of Regulatory triplets with Retrieval-Augmented Generation (RAG) to address these demands. First, agents build and maintain an ontology-free KG by extracting subject--predicate--object (SPO) triplets from regulatory documents and systematically cleaning, normalizing, deduplicating, and updating them. Second, these triplets are embedded and stored along with their corresponding textual sections and metadata in a single enriched vector database, allowing for both graph-based reasoning and efficient information retrieval. Third, an orchestrated agent pipeline leverages triplet-level retrieval for question answering, ensuring high semantic alignment between user queries and the factual "who-did-what-to-whom" core captured by the graph. Our hybrid system outperforms conventional methods in complex regulatory queries, ensuring factual correctness with embedded triplets, enabling traceability through a unified vector database, and enhancing understanding through subgraph visualization, providing a robust foundation for compliance-driven and broader audit-focused applications.

Problem

• Enhancing regulatory QA accuracy with multi-agent KG and RAG integration
• Building ontology-free KG from regulatory documents for precise compliance
• Ensuring factual correctness and traceability in complex regulatory queries

Innovation

• Multi-agent framework integrates KG and RAG
• Ontology-free KG built from regulatory triplets
• Hybrid system combines graph reasoning and retrieval