Écriture des toponymes en français : variations entre normes et usages

This work aims to automatically identify toponyms in text. It concerns natural language processing techniques and deals with toponym French spelling, especially with variations between standards and the way these are applied in a Web corpus. The definitions of toponym are manifold because they are closely related to the fields of study of the relevant disciplines. Because of the abundance of these disciplines, several bodies write reference frameworks about toponym spelling. These standards rely on concepts which fall within the fields of study; they are exposed and discuss: Terminological and ontological resources which are mandatory to automatically identify toponyms in texts are described. A tool is implemented and assessed, which takes into account both reference frameworks and variations found in the corpus. Our results validate the spelling assumptions which were implemented, thus enabling us to draw perspectives.