Monuments from Antiquity and Natural Language Processing (MONUMENTAL)
MonumenTAL is a project that looks at the formulations and expressions used to refer to and distinguish works of figure art from classical Antiquity. These texts, from Antiquity to today, present a variety of titles and thematic terms for monuments, particularly statues. These reflect the formulations and relations to the work that were dominant in the time and the context of their creation, as well as the attempts and principles of formalisation. These terms today pave the way toward to textual data on these cultural heritage works and contribute to the reconstruction of lineages between original visual creations, their copies, and their conceptual equivalents. This project is based on close collaboration between linguists, art historians, and philologists on the study of ancient and contemporary textual sources, for the most part conserved at the French National Library (BnF). It aims to identify the often plural designations of works, to study how they were developed in different contexts, and to bring them together to create a multi-lingual reference that can be operated between texts and digital corpuses of artworks. It aims to have this vocabulary interact with natural learning processes (TAL), and to experiment with deep learning in order to develop tools and processes for the exploration of texts. This research on forms of expression also covers the study of the relationship between the commentator, the work, and his/or her own erudition and readers in different contexts. It also explores the relationship between the artist and the work known only through texts. This knowledge of texts that have been re-discovered or revived will be shared with the public through a selection of digital sources that will be put online, dedicated to these works that have been decisive in artistic creation, and in the writings of collectors, theorists, and scholars in the field of classical art. Finally, the project will enable a cross-fertilization of archaeological vocabulary used in the context of different research and heritage projects and library sciences.
Anne-Violaine SZABADOS , CNRS - ArScAn UMR 7041
Internal cluster partners
- Archéologies et Sciences de l'Antiquité (ArScAn) - UMR 7041
- Bibliothèque nationale de France (BnF)
- Modèles, Dynamiques, Corpus (MoDyCo) - UMR 7114
traitement automatique de la langue-TAL, fouille de textes, patrimoine antique, référentiel iconographie, humanités numériques