Web Tool for Automated Document Formatting Verification

Abstract

The article discusses the algorithms for document compilation check in compliance with standards and regulatory documents. The paper presents various approaches and methods of extraction of structural elements and their properties from PDF, ODT, DOCX documents; reveals machine-learning opportunities in terms of class extraction and further structural elements classification. The machine-learning methods are also used to provide recommendations on possible errors. The article introduces automated document formatting verification service architecture.

Keywords

• catboost parsing document formatting verification pdf odt docx service architecture machine learning nlp software architecture