From Drafting Guideline to Error Detection: Automating Style Checking for Legislative Texts

This paper reports on the development of methods for the automated detection of violations of style guidelines for legislative texts, and their implementation in a prototypical tool. To this aim, the approach of error modelling employed in automated style checkers for technical writing is enhanced to meet the requirements of legislative editing. The paper identifies and discusses the two main sets of challenges that have to be tackled in this process: (i) the provision of domain-specific NLP methods for legislative drafts, and (ii) the concretisation of guidelines for legislative drafting so that they can be assessed by machine. The project focuses on German-language legislative drafting in Switzerland.