Government Departments Urged to Support the Future of Irish Language Technology

24 March 2021

As the official language of the Republic of Ireland and an official language of the European Union, all State-publications, including everything from information booklets to forms to legislation, needs to be made available in the Irish language.  A research team based at the SFI ADAPT Centre at DCU is helping to develop robust machine translation systems for Irish and they are looking for help from the Irish Government to further this research and to ensure the digital future of the Irish language.  ADAPT’s researchers are calling on all 19 Irish Government departments to send them a special file called a Translation Memory File (TMX file), which contains a record of all of the translations between Irish and English that the Departments have paid for from third party translation companies over the years. The campaign is currently underway to coincide with Seachtain na Gaeilge, and the team is asking the Departments to send them their TMX files by Easter Weekend.

Speaking about the campaign, Micheál Ó Conaire, Principal Officer in the Gaeltacht Division of The Department of Tourism, Culture, Arts, Gaeltacht, Sport and Media said: “When you share your Department’s TMX file with DCU researchers you’re helping to ensure the development of Irish language technology now and into the future.  It couldn’t be easier to get your TMX file.  Simply ask the company that does your Department’s translations to send it to you or and then share it with DCU’s researchers before Friday 2nd April”.

Helen McHugh, Irish Language Technology Project Officer at the ADAPT Centre said: “Machine translation systems are data-driven so the more samples of translations they receive, the better the predictions they make and the more accurate they become.  The vast quantities of Government translations between Irish and English could provide the backbone of the next-generation of technology that can support the Irish language and we urge all Departments to get in touch with us.”

Departments that have their TMX file containing Irish language data can send it to the research team by email at [email protected]