Identification of Regulatory Requirements for Business Processes

This technical report examines the identification of regulatory requirements relevant to business processes, addressing the challenges organizations face in ensuring compliance with a growing number of regulatory documents. The study highlights the manual effort required by legal and domain experts to identify relevant documents, which can be cumbersome, especially with frequently changing regulations. The authors compare various methods, including an embedding-based NLP ranking method, a generative AI approach using GPT-4, and a crowdsourced method against the traditional manual method. The evaluation is based on two case studies: an Australian insurance case and a global banking use case. The report discusses the strengths and weaknesses of each method in terms of applicability, automation, transparency, and reproducibility. It also provides recommendations on method combinations that can maximize benefits based on specific characteristics of application scenarios. The findings aim to assist organizations in reducing compliance burdens and enhancing their regulatory assessment processes.