This article presents applied research work that addresses a real need for improvement of the collection, indexing and retrieval of knowledge of cultural heritage. To facilitate the task of fieldwork and to improve the access to knowledge, we designed, prototyped and tested in a real environment a system for automating the cycle of collecting and storing information and knowledge for future processing. It contains several modules, allowing the knowledge acquisition cycle to be completed: registering artwork description using a voice interface, automatic 'translation' of the audio files to text files, linguistic analysis and extraction of information for the Inventory Descriptive System and, finally, automatic construction of a domain ontology according to a predefined conceptual model. This hybrid system combines the techniques of signal and natural language processing and of knowledge modelling. After introducing the process of the acquisition and management of information gathered in the field, we describe the nature of knowledge, the models used, the architecture and the various treatments performed by our system. Finally, we provide some perspectives on future development.
National audienceCet article présente un travail de recherche appliquée répondant à un réel besoin d'amélioration de recueil, d'indexation et de recherche de connaissances dans le domaine du patrimoine culturel. Conçu, modélisé et expérimenté en environnement réel par Stephan du Chateau, Danielle Boulanger, Eunika Mercier-Laurent, le système Simplicius présenté dans cet article permet d'automatiser le cycle de vie d'informations et connaissances pour une recherche future. Il est composé de plusieurs modules permettant le recueil d'informations orales, leur "traduction" automatique vers un fichier texte, l'analyse linguistique et l'extraction d'informations et, enfin, le peuplement semi-automatique d'une ontologie de domaine selon un modèle conceptuel prédéfini
This document presents our work on a definition and experimentation of a voice interface for cultural heritage inventory. This hybrid system includes signal processing, natural language techniques and knowledge modeling for future retrieval. We discuss the first results and give some points on future work.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.