Titre : Computational Characterization and Taxonomy of Patent Writing Styles: Towards Stylometric Automation of Technical Drafting 

Auteurs : Ghislain DEMONDA
, Sébastien LABORIE
, Christian SALLABERRY
, Nathalie VALLES-PARLANGEAU,  

Revue : Open Journal in Information Systems Engineering 

Numéro : Special Issue INFORSID 2025 

Volume : 6 

Date : 2026/04/22 

DOI : 10.21494/ISTE.OP.2026.1450 

ISSN : 2634-1468 

Résumé : In the field of intellectual property, patents are essential technical and legal documents whose drafting requires expertise that combines technical, legal, and linguistic skills. Patent drafting styles vary considerably depending on technological domains, jurisdictions, and protection strategies. This article proposes the design of SCASB (System for Stylometric Characterization and Automation of Patents), an approach that, for the first time, unifies the technical, legal, and stylistic dimensions within a coherent computational framework. We propose a two-dimensional taxonomy of patent analysis approaches (automatic computational methods for document analysis × level of analytical granularity) and highlight current shortcomings. Our system builds upon the rapid advances in artificial intelligence technologies, particularly natural language processing (NLP). This work opens the path toward intelligent automation of technical drafting that accounts for the stylistic nuances specific to each jurisdiction and protection strategy. 

Éditeur : ISTE OpenScience