Project

Back to overview

Multilingual Translation Model Based on Object-Oriented design: improvements and enhancements

Applicant Wehrli Eric
Number 130634
Funding scheme Project funding (Div. I-III)
Research institution Département de Linguistique Française Faculté des Lettres Université de Genève
Institution of higher education University of Geneva - GE
Main discipline Other languages and literature
Start/End 01.09.2010 - 31.07.2012
Approved amount 118'336.00
Show all

All Disciplines (2)

Discipline
Other languages and literature
Information Technology

Keywords (8)

Machine translation; Syntax; Multilingual model; Chomskyan linguistics; object-oriented design; collocations; generative linguistics; collocation

Lay Summary (English)

Lead
Lay summary
This project is a continuation of our research on the development of a multilingual translation model. Based on (Chomskyan) generative grammar for its linguistic aspects and on an object-oriented design for its implementation, this model aims at drastically simplifying the task of adding language-pairs in a truly multilingual MT system.The proposed enhancements will concern primarily the treatment of complex collocations, the question of pronominal reference (intrasentential as well as intersentential). In addition, significant improvements will be brought to the crucial task of automatic creation of bilingual lexical databases by transitivity. Several language pairs will be added to the system, to cover all the possible pairs for 5 of the major European languages (English, French, German, Italian and Spanish). The system will be evaluated as part of the WMT machine translation evaluation campaign.
Direct link to Lay Summary Last update: 21.02.2013

Responsible applicant and co-applicants

Employees

Publications

Publication
Les lexiques bilingues pour les traducteurs automatiques ou l’apport de la lexicographie computationnelle
Russo Lorenza (2012), Les lexiques bilingues pour les traducteurs automatiques ou l’apport de la lexicographie computationnelle, in Dotoli Boccuzzi Lo Nostro (ed.), Schena-A. Baudry et Cie, Fasano (Italie), Paris (France), 197-212.
Italian and Spanish Null Subjects. A Case Study Evaluation in an MT Perspective
Russo Lorenza, Loaiciga Sharid, Gulati Asheesh (2012), Italian and Spanish Null Subjects. A Case Study Evaluation in an MT Perspective, in Lrec 2012, Turquie.
Improving Machine Translation of Null Subjects in Italian and Spanish
Russo Lorenza, Loaiciga Sharid, Gulati Asheesh (2012), Improving Machine Translation of Null Subjects in Italian and Spanish, in EACL 2012, France.
Étude inter-langues de la distribution et des ambiguïtés syntaxiques des pronoms
Russo Lorenza, Scherrer Yves, Goldman Jean-Philippe, Loaiciga Sharid, Nerima Luka, Wehrli Eric (2011), Étude inter-langues de la distribution et des ambiguïtés syntaxiques des pronoms, in TALN 2011, France.
La traduction automatique des pronoms. Problèmes et perspectives
Scherrer Yves, Russo Lorenza, Goldman Jena Philippe, Loaiciga Sharid, Nerima Luka, Wehrli Eric (2011), La traduction automatique des pronoms. Problèmes et perspectives, in TALN 2011, France.
La traduction automatique des séquences clitiques dans un traducteur à base de règles
Russo Lorenza, Wehrli Eric (2011), La traduction automatique des séquences clitiques dans un traducteur à base de règles, in TALN 2011, France.
La traduction automatique entre langues proches : le cas de l'italien et du français
Russo Lorenza (2011), La traduction automatique entre langues proches : le cas de l'italien et du français, in Actes des XXIV Journées de Linguistique (JDL), Québec, Canada.
The automatic translation of clitic pronouns: the Its-2 system
Russo Lorenza (2010), The automatic translation of clitic pronouns: the Its-2 system, in GG@G - Generative Grammar in Geneva, 6, 203-220.
La traduction automatique des pronoms clitiques. Quelle approche pour quels résultats ?
Russo Lorenza (2010), La traduction automatique des pronoms clitiques. Quelle approche pour quels résultats ?, in TALN 2010, Montréal.
A Recursive Treatment of Collocations
Nerima Luka, Wehrli Eric, Seretan Violeta (2010), A Recursive Treatment of Collocations, in LREC 2010, Malta.
Pour une analyse linguistique comparative italien-français en vue d’une traduction automatique
Russo Lorenza, Pour une analyse linguistique comparative italien-français en vue d’une traduction automatique, Publif@um, Italie.

Scientific events

Active participation

Title Type of contribution Title of article or contribution Date Place Persons involved
LREC 2012 21.05.2012 Istanbul, Turquie
EACL 2012 16.04.2012 Avignon, France
Journée ``Recherches doctoraux en français en Italie -- II édition: Nos voix/voies de recherche'' 14.11.2011 Turin, Italie
Le Dictionnaire Bilingue. Tradition et innovation - Neuvièmes Journées des Dictionnaires 26.09.2011 Bari, Italie
Taln 2011 27.06.2011 Montpellier, France
Workshop on Automatic Language Analysis: Fips and Freeling resources 24.01.2011 Barcelone, Espagne
Ecole d'hiver TRIPLE 2011 17.01.2011 Rome, Italie
6e Journées de Linguistique Suisse 06.12.2010 Neuchatel, Suisse
TALN 2010 19.07.2010 Montréal, Canada
LREC 2010 24.05.2010 Malte
XXIVes Journées de Linguistique (JDL) 22.03.2010 Laval, Canada


Associated projects

Number Title Start Funding scheme
113864 Multilingual Translation Model Based on Object-Oriented 01.10.2006 Project funding (Div. I-III)
113864 Multilingual Translation Model Based on Object-Oriented 01.10.2006 Project funding (Div. I-III)

Abstract

This project is a continuation of our research on the development of a multilingual translation model. Based on (Chomskyan) generative grammar for its linguistic aspects and on an object-oriented design for itsimplementation, this model aims at drastically simplifying the task of adding language-pairs in a truly multilingual MT system.The proposed enhancements will concern primarily the treatment of complex collocations, the question of pronominal reference (intrasentential as well as intersentential). In addition, significant improvements will be brought to the crucial task of automatic creation of bilingual lexical databases by transitivity. Several language pairs will be added to the system,to cover all the possible pairs for 5 of the major European languages (English, French, German, Italian and Spanish). The system will be evaluated as part of the WMT machine translation evaluation campaign.
-