This paper describes our work on building a trilingual parallel treebank. We have annotated constituent structure trees from three text genres (a philosophy novel, economy reports and a technical user manual). Our parallel treebank includes word and phrase alignments. The alignment information was manually checked using a graphical tool that allows the annotator to view a pair of trees from par...

Transliterations and cognates have been shown to be useful in the case of bilingual extraction from parallel corpora. Observation of transliterations in a trilingual English, French and Japanese specialised comparable corpus reveals evidences that they are likely to be used with comparable corpora too, since they are an important and relevant part of the common vocabulary, but they also yield l...

This paper presents the Copa 2014 FrameNet Brasil software (C-14/FN-Br): a frame-based trilingual electronic dictionary covering the domains of Football, Tourism and the World Cup. The dictionary relies on the infrastructure of FrameNet and is meant to be used by tourists, journalists and the staff involved in receiving foreign visitors. Vocabulary from the three domains is made available in En...

The aim of this research was to investigate the language functions of trilinguals that live in Greece. This study consisted of personal interviews conducted at a State Evening School in Volos in January 2010 and the participants were between 19 and 24 years old. They were all male and they came from immigrant families of Albanian origin. The study aimed to locate the instances that each one of ...

The JHU HLTCOE participated in the Cold Start and the Trilingual Entity Linking and Discovery tasks of the 2015 Text Analysis Conference Knowledge Base Population evaluation. For our fourth year of participation in Cold Start we continued our research with the KELVIN system. We submitted experimental variants that explore use of linking to Freebase and adding additional relations. This is our f...

(for dissemination) We describe the experimental results using the baseline speech-to-speech translation systems created in D4.3 and compare them to an enhanced translation system taking different language resources into account. Experiments were performed on the trilingual corpus (English, Spanish, Catalan) built within the project in WP5. This corpus consists of spontaneous dialogues in the d...