MaTIAS - Machine Translation to Inform Asylum Seekers



This project aims to develop a multilingual messaging system for asylum reception facilities in Belgium. The messaging system will allow staff from asylum reception centres to convey practical messages and instructions to centre residents in their own language.

We will create an open source prototype that allows simple text messages to be translated and forwarded through an existing messaging system (e.g. SMS, Whatsapp, Signal, Telegram). This prototype will be based on a combination of a Fedasil-specific or context-specific translation memory on the one hand (i.e. a database of text messages and professional translations in the supported languages) and machine translation on the other. The prototype is built according to a two-step logic: (1) translations are generated based on the context-specific translation memory; (2) if the text message has not been previously translated or the target language is not available, the messages are generated via machine translation .

This interdisciplinary project combines methods from ethnography and translation technology. It is a close collobaration between two research groups: MULTIPLES (July De Wilde, Katrijn Maryns and Ella van Hest) and LT3 (Lieve Macken, Arda Tezcan and Michael Lumingu). The project is funded by the European Union.




Postdoctorale medewerker(s)

Technische medewerker(s)