eContent Localization Resources for Translator Training
This website contains the contribution of the University of Geneva to the eCoLoRe project. The purpose of this project is to create shareable and renewable e-content localisation resources to support translator training. These resources include sample texts, pedagogic scenarios, and other instructional material necessary for this type of training.
The contribution of the University of Geneva is divided into three modules: Localization, Machine Translation, and Translation Memory Systems.
TransWeb Express: a sample localisation project
Machine Translation (MT) is a technology belonging to computational linguistics, which allows specialized computer programs to translate texts from one human language into another automatically. Although the technology available today does not yet provide high quality translations, it is possible to improve the results substantially by personalizing the MT systems and by applying certain rules to the source texts before they are translated. This web site provides material that illustrates the process of MT improvement, describing each step of the process with theoretical explanations and practical examples. It also provides suggestions for introductory MT exercices.
Translation Memory Systems
A machine assisted translation system, or a translation memory system, is a software functioning as a translation aid, as opposed to a fully automatic machine translation system. In this section, we present a course on the use of machine assisted translation systems. The course is divided in two major parts: a theoretical part, Base théorique, which introduces the basic aspects of translation memories and the specific functionalities of the systems, and a practical part, Travaux pratiques, presenting exercises on how a memory is built by translated texts, aligned texts, or imported memories.