LIST represents Luxembourgish in European Language Equality project

Funded by the European Union, the European Language Equality (ELE) project is a consortium of 52 partners from all over the bloc with an objective of developing an agenda and a roadmap for achieving full digital language equality in Europe by 2030.

Fighting the cause for Luxembourg and the Luxembourgish language is the Luxembourg Institute of Science and Technology (LIST) as the only country representative within the consortium.

Heading up the project at LIST is Dimitra Anastasiou, a data science and analytics researcher in the IT for Innovations Services department. She explained her role in ELE. “What I have done is to collect metadata about anything that has to do with Luxembourgish, meaning data resources, text documents either bilingual or monolingual, dictionaries and glossaries etc, as well as any tools and applications that promote Luxembourgish or are written in Luxembourgish”.

In January Dimitra submitted a report regarding the metadata collected that included the landscape of language technology for Luxembourgish, including available tools and resources as well as projects, initiatives, and stakeholders. This report was also sent to the University of Luxembourg who study the Luxembourgish language, to the government’s Luxembourgish Language Commissioner Marc Barthelemy, and to “Zenter fir d'Lëtzebuerger Sprooch” as the report makes frequent reference to Luxembourgish spelling and ZLS is the centre responsible for this.

But promoting Luxembourgish is not as easy as promoting the country’s other official languages, French and German as Dimitra explained. “It is difficult because Luxembourgish is an under resourced language so there’s a lot of data lacking. I would say the reason for that is because Luxembourg is so multilingual and multinational with percentage of foreigners reaching almost 50%. Everyone speaks French, German or English, so people communicate in another language apart from Luxembourgish because they don’t need to,” she outlined.

Indeed, the Grand Duchy’s national language is Luxembourgish, while French is the legislative language. French, German and Luxembourgish are the three administrative and judicial languages.

“The second reason it is difficult is because it is not currently an official EU language, which means that official EU documents are not translated into Luxembourgish,” said Dimitra.

However, should Luxembourgish eventually reach official EU status it will have a huge impact on language technology as a huge amount of data will be created. “All the EU websites will be in Luxembourgish and through data crawling a lot of data will be processed,” said Dimitra. “However, this entails a high investment both in terms of cost and time to translate all EU official documents into Luxembourgish, which should be taken into account by many stakeholders.”

What is the ultimate aim of ELE?

The main aim is to reach digital equality for most European languages which also includes many variants. This should mean that the under resourced languages like Luxembourgish have better representation on the internet and the government.

The ELE project is a direct response to the resolution “Language equality in the digital age”, which was passed by the European Parliament in a landslide vote in September 2018. It takes into account more than 70 languages and has received funding from the European Union.

The involvement of Luxembourgish in ELE therefore begs the question: Could it be a step towards Luxembourgish becoming an official EU language? “Absolutely!” said Dimitra, “it is a long-term vision of institutions, organisations and the government to promote Luxembourgish and have it in a digital well-represented form, and we have seen the recent arrival of language technology projects, like the Chatbot. It shows that LIST has received funding in the past few years regarding language technology, so I am very optimistic about the future as one project brings another,” she said before elaborating, “I would like to continue working on natural language processing as I see it as a very promising domain in many disciplines like bio and health informatics, banking, entertainment and communication”.

But working with a very small language poses a big challenge. While support in French, German or Spanish is reasonably easy as resources are rife in all aspects, for Luxembourgish so much is missing and, in many cases, everything has to be created from scratch. As written Luxembourgish is not taught in general in schools and is mainly an oral language, even spelling can be not consistent hence LIST’s consultation with ZLS in the ELE project.

The ELE project at LIST, and indeed across Europe, currently continues until June 2023 with a possible follow-up thereafter.

LIST is also a National Competence Centre in a sister project of ELE, the European Language Grid (ELG), whose aim is to create a powerful and scalable language technology platform using a grid architecture with thousands of datasets, and hundreds of language technology services.