At the Dutch Tax authority, the Law text is evidently the starting point of our daily operations including the implementation in software applications and systems. The Law text is, however, not usable at the operational level and needs to be interpreted and supplemented.

To deal with this issue, we have introduced a semi-automated process that starts at the Law text and generates the necessary content that is required at the next level while keeping the link to the original sources. The final product of this process is a generated software application that executes the Law without any manual programming.

Learning Objectives:

• Automate text extraction and linking of Law text to business rules
• Create a natural language pattern that fits Tax related operational rules
• Make natural language rules readily executable