Automated Transformation of Financial Regulations into SBVR: An Ontology-Based and Natural Language Processing Approach

Institutions face regulatory complexity, where traditional Governance, Risk, and Compliance (GRC) approaches are often reactive and prone to errors. Legal language also introduces ambiguity into compliance traceability. Recent advances in Natural Language Processing (NLP) offer potential solutions for interpreting regulations. This study introduces CFR2SBVR, an automated method for transforming natural-language financial regulations into structured vocabularies and rules aligned with the Semantics of Business Vocabulary and Business Rules (SBVR) standard, thereby making it easier to trace regulatory requirements to business operations. The method combines NLP techniques, Large Language Models (LLMs), and ontology mappings, such as the Financial Industry Business Ontology (FIBO), to extract, classify, and formalize regulatory content from the U.S. Code of Federal Regulations (CFR). Grounded in the Design Science Research (DSR) framework (Wieringa, 2014), this study develops and validates artifacts through multiple checkpoints, ensuring traceability between original texts and transformed statements. The evaluation employed semantic similarity metrics (SemScore and LLM-as-a-Judge) and SME validation, achieving an average accuracy of 0.85 or higher and an overall accuracy of 95.81%. Compared to deterministic and hybrid approaches, CFR2SBVR demonstrates greater adaptability and scalability, particularly in managing ambiguity and responding to changing regulatory requirements. All datasets, source code, and results are publicly available to promote reproducibility and further research in GRC.