GraphCompliance: Aligning Policy and Context Graphs for LLM-Based Regulatory Compliance

2025-10-30

AI Summary

This work addresses the challenge of aligning large language models (LLMs) with structured regulatory texts—particularly when reasoning over unstructured, context-rich operational scenarios in regulatory compliance assessment. We propose a dual-graph alignment framework: (1) modeling regulations as a policy graph encoding normative logic and cross-references, and (2) representing runtime contexts as a context graph composed of SAO (Subject–Action–Object)-based entity-relation triples. Alignment is achieved via graph-structured semantic parsing, augmented by a judge LLM and retrieval-augmented generation (RAG) for interpretable, structured reasoning. Evaluated on 300 real-world GDPR compliance scenarios, our method achieves a micro-F1 improvement of 4.1–7.2 percentage points over LLM-only and RAG-only baselines, with higher recall and lower false-positive rates. Ablation and attribution analyses confirm the critical role of dual-graph representation and alignment in enhancing both accuracy and explainability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• 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
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Compliance at web scale poses practical challenges: each request may require a regulatory assessment. Regulatory texts (e.g., the General Data Protection Regulation, GDPR) are cross-referential and normative, while runtime contexts are expressed in unstructured natural language. This setting motivates us to align semantic information in unstructured text with the structured, normative elements of regulations. To this end, we introduce GraphCompliance, a framework that represents regulatory texts as a Policy Graph and runtime contexts as a Context Graph, and aligns them. In this formulation, the policy graph encodes normative structure and cross-references, whereas the context graph formalizes events as subject-action-object (SAO) and entity-relation triples. This alignment anchors the reasoning of a judge large language model (LLM) in structured information and helps reduce the burden of regulatory interpretation and event parsing, enabling a focus on the core reasoning step. In experiments on 300 GDPR-derived real-world scenarios spanning five evaluation tasks, GraphCompliance yields 4.1-7.2 percentage points (pp) higher micro-F1 than LLM-only and RAG baselines, with fewer under- and over-predictions, resulting in higher recall and lower false positive rates. Ablation studies indicate contributions from each graph component, suggesting that structured representations and a judge LLM are complementary for normative reasoning.

Problem

• Aligning unstructured text with structured regulatory requirements
• Reducing interpretation burden for regulatory compliance assessments
• Improving accuracy in GDPR scenario evaluations through graph alignment

Innovation

• Policy Graph represents regulatory texts structure
• Context Graph formalizes events as triples
• Graph alignment anchors LLM reasoning in structure