The goal of the project was the development of an interactive corpus-based translation drafting tool. The system is based upon sophisticated pattern matching techniques (both linguistic and statistical) used to identify the longest coherent part of source text which has already been translated and has been appropriately stored in a text database. This part was then presented to the user for confirmation, adoption or correction. The Translearn tools were developed for four languages: Greek, English, French and Portuguese. The results of the project were quite positive.
These translation tools proved very important for the work of the translator. They decrease the variety of translation where it is not needed, while they make the most out of good quality translations. Finally the translator's productivity is increased substantially.