Explainable Compliance Detection with Multi-Hop Natural Language Inference on Assurance Case Structure
This addresses the challenge of ensuring compliance with regulations like GDPR for complex systems, offering an automated and explainable solution, though it appears incremental as it builds on existing NLI and LLM methods.
The paper tackles the problem of automating regulatory compliance detection for complex systems by proposing a Natural Language Inference (NLI) approach called EXCLAIM, which formulates assurance cases as multi-hop inference for explainable detection and uses LLMs to generate assurance cases, demonstrating effectiveness with GDPR requirements as a case study.
Ensuring complex systems meet regulations typically requires checking the validity of assurance cases through a claim-argument-evidence framework. Some challenges in this process include the complicated nature of legal and technical texts, the need for model explanations, and limited access to assurance case data. We propose a compliance detection approach based on Natural Language Inference (NLI): EXplainable CompLiance detection with Argumentative Inference of Multi-hop reasoning (EXCLAIM). We formulate the claim-argument-evidence structure of an assurance case as a multi-hop inference for explainable and traceable compliance detection. We address the limited number of assurance cases by generating them using large language models (LLMs). We introduce metrics that measure the coverage and structural consistency. We demonstrate the effectiveness of the generated assurance case from GDPR requirements in a multi-hop inference task as a case study. Our results highlight the potential of NLI-based approaches in automating the regulatory compliance process.