MaTIAS – Machine Translation to Inform Asylum Seekers

With this project, we aim to develop an open source prototype of a multilingual messaging system that can translate and forward simple messages via an existing messaging application (e.g. WhatsApp). The messaging system will be developed in cooperation with Fedasil, the Belgian federal agency for the reception of asylum seekers. The messaging system will allow asylum reception centre staff to convey practical messages and instructions to centre residents (e.g. inform centre residents about an activity or technical problem), including for so-called ‘language of lesser diffusion’. For translation, the messaging system relies on both a context-specific translation memory and on machine translation.