Implementing regulatory documents is a recurring, mostly manual and time-consuming task for companies. To establish and ensure regulatory compliance, constraints need to be extracted from the documents and integrated into process models capturing existing operational practices. Since regulatory documents and processes are subject to frequent change, the constant comparison between both is mandatory. Additionally, new regulations must be integrated and checked against existing process models. To address these challenges, we provide an approach that uses natural language processing to automatically support compliance assessment between regulatory documents and process model repositories. The outcome is a pairwise matching between parts of a regulatory document and process models from a repository. This matching can be used to either determine the coverage of regulations by a process model or to guide compliance assessment by ranking models based on their fitness and cost. The approach is implemented and applied in two real-world case studies: one from the energy domain and the other based on the General Data Protection Regulation.

Top