spoken Language UNderstanding in multilinguAl communication systems (LUNA)

SIXTH FRAMEWORK PROGRAMME PRIORITY 2 “INFORMATION SOCIETY TECHNOLOGY”
SPECIFIC TARGETED RESEARCH OR INNOVATION PROJECT

The main goal of the LUNA project is the creation of a robust natural SLU toolkit for multilingual dialogue services, able to carry on human-computer communication with a good degree of user satisfaction. It is evident from the evolution of research in the field that more realistic (and complex) structural models are needed even if they will be used in a framework of statistical decision theory. Effective structural models are language dependent and the European researchers, especially those who are recognized world leaders in the field, are in a better position to develop and test these models, especially for the European languages.LUNA will attempt to improve interpretation accuracy by exploiting the potential of integrating ASR and SLU technologies. From a technological point of view, LUNA’s objectives include the investigation of new methods, algorithms and tools for the fast development of robust SLU components for multilingual telecom services.
To this aim, LUNA will address a set of challenging scientific problems, by focusing on five scientific objectives:

•     Language Modelling for Speech Understanding

•     Semantic Modelling for Speech Understanding

•     Automatic Learning (including Active and On-Line Learning)

•     Robustness issues for SLU

•     Multilingual portability of SLU components

Consortium partners:

•     Loquendo

•     University of Avignon

•     University of Trento

•     CSI Piemonte

•     RWTH Aachen

•     France Telecom Division R&D

•     Japanese Institute of Information Technology

•     Instytut Podstaw Informatyki PAN