id
stringlengths
18
23
title
stringlengths
13
211
abstract
stringlengths
93
2.68k
keyphrases
sequence
prmu
sequence
translation
sequence
taln-2010-demo-007
Traitement automatique des langues des signes : le projet Dicta-Sign, des corpus aux applications
Cet article présente Dicta-Sign, un projet de recherche sur le traitement automatique des langues des signes (LS), qui aborde un grand nombre de questions de recherche : linguistique de corpus, modélisation linguistique, reconnaissance et génération automatique. L’objectif de ce projet est de réaliser trois applications prototypes destinées aux usagers sourds : un traducteur de termes de LS à LS, un outil de recherche par l’exemple et un Wiki en LS. Pour cela, quatre corpus comparables de cinq heures de dialogue seront produits et analysés. De plus, des avancées significatives sont attendues dans le domaine des outils d’annotation. Dans ce projet, le LIMSI est en charge de l’élaboration des modèles linguistiques et participe aux aspects corpus et génération automatique. Nous nous proposons d’illustrer l’état d’avancement de Dicta-Sign au travers de vidéos extraites du corpus et de démonstrations des outils de traitement et de génération d’animations de signeur virtuel.
[ "Langue des signes", "génération automatique", "corpus vidéo comparables", "reconnaissance automatique" ]
[ "P", "P", "M", "R" ]
[ "", "This paper presents Dicta-Sign, a research project related to sign language (SL) processing. It covers numerous research topics: corpus linguistics, linguistic modelling, automatic recognition and generation. The aim of this project is to design three proof-of-concept end user applications: an SL-to-SL term translator, a search-by-example tool, and a SL wiki. For that, four comparable corpora of five hours will be produced and analysed. Aside from these applications, we also expect major improvements to be integrated to annotation tools. In this project, LIMSI is in charge of the linguistic modelling, and participates in building the corpus and in the generation efforts. We propose to illustrate the current work with excerpts from the corpus and demonstrations of the processing and generation tools.", "Sign languages, comparable video corpora, automatic recognition, automatic generation" ]
taln-2010-demo-008
FipsColor : grammaire en couleur interactive pour l’apprentissage du français
L'analyseur multilingue FiPS permet de transformer une phrase en une structure syntaxique riche et accompagnée d'informations lexicales, grammaticales et thématiques. On décrit ici une application qui adapte les structures en constituants de l’analyseur FiPS à une nomenclature grammaticale permettant la représentation en couleur. Cette application interactive et disponible en ligne (http://latl.unige.ch/fipscolor) peut être utilisée librement par les enseignants et élèves de primaire.
[ "analyse syntaxique", "grammaire générative", "services web", "tei" ]
[ "M", "M", "U", "U" ]
[ "", "The FiPS parser analyzes a sentence into a syntactic structure reflecting lexical, grammatical and thematic information. The present paper describes the adaptation of the structures in terms of constituents as existent in FiPS to a grammatical annotation, as well as its coloured representation. This online interactive application (available at http://latl.unige.ch/fipscolor) can be freely used by teachers and pupils of primary education.", "chart parser, generative grammar, web services, tei" ]
taln-2010-demo-009
Des cartes dialectologiques numérisées pour le TALN
Cette démonstration présente une interface web pour des données numérisées de l’atlas linguistique de la Suisse allemande. Nous présentons d’abord l’intégration des données brutes et des données interpolées de l’atlas dans une interface basée sur Google Maps. Ensuite, nous montrons des prototypes de systèmes de traduction automatique et d’identification de dialectes qui s’appuient sur ces données dialectologiques numérisées.
[ "Dialectologie", "atlas linguistique", "traduction automatique", "identification de dialectes" ]
[ "P", "P", "P", "P" ]
[ "", "This demonstration presents a web interface for digitized data of the linguistic atlas of German-speaking Switzerland. First, we present the integration of raw and interpolated atlas data with an interface based on Google Maps. Then, we show prototypes of machine translation and dialect identification systems which rely on the digitized dialectological data.", "Dialectology, linguistic atlas, machine translation, dialect identification" ]
taln-2010-demo-010
Text-it /Voice-it Une application mobile de normalisation des SMS
Cet article présente Text-it / Voice-it, une application de normalisation des SMS pour téléphone mobile. L’application permet d’envoyer et de recevoir des SMS normalisés, et offre le choix entre un résultat textuel (Text-it) et vocal (Voice-it).
[ "application", "normalisation", "SMS", "plugin", "serveur" ]
[ "P", "P", "P", "U", "U" ]
[ "", "This paper presents Text-it / Voice-it, an application that makes it possible to normalize text messages directly from mobile phones. The application allows the user to both send and receive normalized text messages, and gives the choice between a textual (Text-it) and a vocal (Voice-it) result.", "Text messages, normalization, application, plugin, server" ]
taln-2010-demo-012
MeTAE : Plate-forme d’annotation automatique et d’exploration sémantiques pour le domaine médical
Nous présentons une plate-forme d’annotation sémantique et d’exploration de textes médicaux, appelée « MeTAE ». Le processus d’annotation automatique comporte une première étape de reconnaissance des entités médicales présentes dans les textes suivie d’une étape d’identification des relations sémantiques qui les relient. Cette identification se fonde sur des patrons linguistiques construits manuellement pour chaque type de relation. MeTAE génère des annotations RDF à partir des informations extraites et offre une interface d’exploration des textes annotés avec des requêtes sous forme de formulaire. La plate-forme peut être utilisée pour analyser sémantiquement les textes médicaux ou interroger la base d’annotation disponible pour avoir une/des réponses à une requête donnée (e.g. « ?X prévient maladie d’Alzheimer », équivalent à la question « comment prévenir la maladie d’Alzheimer ? »). Cette application peut être la base d’un système de questions-réponses pour le domaine médical.
[ "domaine médical", "Annotation sémantique", "interrogation sémantique" ]
[ "P", "P", "R" ]
[ "", "This paper presents MeTAE, a platform for semantic annotation and exploration of medical texts. The annotation process encompasses medical entity recognition and semantic relationship identification between the retrieved entities. This identification is based on linguistic patterns constructed manually for each type of relation. MeTAE generates RDF annotations from the extracted information and allows semantic exploration of the annotated texts through a form-based interface. The platform can be used to semantically analyze medical texts or to explore the available annotation base through structured queries (e.g. “?X Prevents Alzheimer’s disease” for its natural-language equivalent: “how to prevent Alzheimer’s disease?”). MeTAE can be a basis for a medical question-answering system.", "Semantic annotation, semantic querying, medical domain" ]
taln-2010-demo-013
LEOPAR, un analyseur syntaxique pour les grammaires d’interaction
Nous présentons ici l’analyseur syntaxique LEOPAR basé sur les grammaires d’interaction ainsi que d’autres outils utiles pour notre chaîne de traitement syntaxique.
[ "grammaires d’interaction", "Analyse syntaxique", "polarités" ]
[ "P", "M", "U" ]
[ "", "We present the parser LEOPAR which is based on the Interaction Grammars formalism. We present also other tools used in our framework for parsing.", "Parsing, Interaction Grammars, polarities" ]
taln-2010-demo-014
TransSearch : un moteur de recherche de traductions
Malgré les nombreuses études visant à améliorer la traduction automatique, la traduction assistée par ordinateur reste la solution préférée des traducteurs lorsqu’une sortie de qualité est recherchée. Cette démonstration vise à présenter le moteur de recherche de traductions TransSearch. Cetteapplication commerciale, accessible sur leWeb, repose d’une part sur l’exploitation d’un bitexte aligné au niveau des phrases, et d’autre part sur des modèles statistiques d’alignement de mots.
[ "alignement de mots", "Traduction automatique statistique", "repérage de traductions", "requêtes linguistiques" ]
[ "P", "R", "M", "U" ]
[ "", "Despite the impressive amount of studies devoted to improving the state of the art of machine translation, computer assisted translation tools remain the preferred solution of human translators when publication quality is of concern. This demonstration presents the translation search engine TransSearch. This web-based commercial application relies on a sentence-aligned bitext and a statistical word alignment techniques.", "Statistical machine translation, translation spotting, word alignment, linguistic queries" ]