LC-STAR

Launched officially in 1st of February 2002, is a project focusing on creating language resources (LR) for transferring Speech-to-Speech Translation (SST) components and thus improving human-to-human and man-machine communication in multilingual environments.

LC-STAR is funded by the European Commision in the scope of the IST programme. The goal of the project is to create lexica for 13 languages and text corpora for 3 languages and to create a demonstrator translating within 3 languages. The specification of lexica and text corpora will be done first.

The work is performed by two parallel running tracks of which Track I (duration 2 years) will concentrate on specification and creation of large word lists and lexica while the Track II (duration 3 years) will concentrate among others on investigation, specification and creation of LR needed in speech centered translation technologies and also building a demonstrator for speech to speech translation. Project details can be viewed on the page called Project Overview.

The major goal of the LC-Star data collection is to enable the project's participants to design and develop multilingual interactive communication services. The set of possible applications using the LC-Star databases will be speech-based. In order to develop such applications, spoken language resources first must be created that adequately cover various aspects of speech variation.