Automatic linguistic analysis of annotated text to improve standards development

The draft Standards in progress, when accompanied by reviewers' comments, represent a clear example of annotated textual data and are therefore attractive for "supervised machine learning" analyses. This article describes an ongoing experiment in which a collection of annotated drafts is analyzed using automatic linguistic analysis techniques to identify effective solutions for improving the Standards creation process. The goal is to predict the parts of the standard that are prone to comments. The results achieved so far are presented and discussed along with new possible directions to take.