Dr. Teresa Lynn:

I refer to the topic of corporate social responsibility. Irish is our first official language. It is considered a minority language in terms of the number of speakers who speak it. There are a couple of reasons why technology is behind. First, because we have a small population of Irish speakers, large technology companies do not see it as a financial investment. There is no return on investment. That is a worldwide problem. There are various ways around trying to convince some of these technology companies that we have a digital presence. First, we could increase the size of our Wikipedia presence and increase the number of people who are tweeting. We could try to get more people in the digital world and then those companies might realise Irish is in demand in a digital world and want to support these people. It usually comes down to numbers like that.

We are not the only language that is being left behind when it comes to these large tech companies. There are some companies which have supported us and there are some tools available to us. I will not name any names but most of them have done it through voluntary translators. They have had volunteers translate the software; they have not paid for it. They pay for the software engineers’ time, but they do not actually pay for the translations. My argument, and that in the report, is that because we are host to the main headquarters for a lot of these companies, we could at least put some pressure on from the perspective of corporate social responsibility.

In terms of work with the Department of Tourism, Culture, Arts, Gaeltacht, Sport and Media with special responsibility for the Gaeltacht, I see Deputy McHugh here who has helped and supported us in the past. I received funding from the Department for over seven years for a research programme which funded PhD and masters students and we had a few interns come through as well. However, there is very little you can do on a small project, that is not long term, large scale and with a wider support. The wider support needs to be from the university itself and from the science groups and so on. The Department has been fantastic with the little knowledge it has of artificial intelligence, AI. That is said with no offence to the Department. It has really been fantastic because it has seen the need for the work. In the past, we have built machine translation systems for use within the Department with responsibility for the Gaeltacht and have been working with the European Commission to try to improve machine translation and the e-translation system with is used by the Directorate-General for Translation, DGT, in Europe by making a co-ordinated effort of collecting translated documents from across public administration. These are called parallel texts in which English and Irish has been aligned and professionally translated to train these machine translation systems. As it is taxpayers’ money which pays for the translation, much of that data is available to us under the open data directive. That is why I mentioned that directive earlier. In this world of machine learning and AI, many of these systems are reliant on data to build them.

We need a more co-ordinated effort in trying to collect this data and in having the experts who are able to engineer the data to build these systems.

We work closely with the Department in writing the digital plan, which we launched in December with the Minister of State, Deputy Chambers. Out of that the ongoing discussions are around what we can do and what type of tools we can build, not for research purposes alone but for practicality. For example, there might be a writing assistant that could help parents communicate with teachers in a Gaelscoil. It could be something as simple as that or maybe making the machine translation systems more widely available across different Government Departments and not just within the Department of Tourism, Culture, Arts, Gaeltacht, Sport and Media. These have to be tailor-made; Google translate is no good. We need to tailor-make these systems to the genre of text that needs to be translated, perhaps for the language officers in various Departments around the country, universities or even within schools. The idea is not to have a machine translation system that would remove a human; it always needs to be post-edited and checked by the human but it reduces the cost of translation, if nothing else, and ensures consistency. It also allows people to do other parts of their jobs and not be bogged down with translation. We have various conversations ongoing, resulting from the publication of the digital plan, as to what we could do next.