Traitement informatique des entités nommées dans un corpus de données foncières en latin : méthodologie et application

An interdisciplinary approach based on the analysis of written, iconographic and archeological data using tools and methods derived from the humanities and social sciences as well, such as geomatics and information sciences, enables to reconstruct the typology and the topology of medieval urban spaces in a broad geo-historical framework. Our research focuses on Avignon between the 13th and 15th centuries and is primarily based on an analysis of homogeneous textual sources written in Latin, using a named entities recognition program. These entities, combined with other types of sources, form a corpus referring to data which have, most of the time, a location. This is why we choose to formalize a geographical data model and a data base, to build a Geographical Information System. This tool allows us to compare, query and geovisualize a wide variety of data in order to analyze the urban space evolution and structure in Avignon during this middle age historical period