Dear reader,

More than 20,000 people from all over Europe have already participated in our survey on the digital support for European languages – but we are aiming higher! In order to include as many voices as possible in our strategic agenda and roadmap towards Digital Language Equality (DLE) by 2030, we have decided to keep the survey open longer. Please participate now and find out more in the ELE section of the newsletter, which also features a great example of the importance of DLE in the shape of an article on tackling the digital gaps for the Irish language.

Meanwhile, the ELG initiative continues its effort to make ELG’s machine translation models for Ukrainian available through a single, publicly accessible interface. The current state is described in the ELG section, followed by first details on this year’s META-FORUM 2022, which is only two months away.

Finally, we present Barcelona Supercomputing Center (BSC), partner in the ELE project and one of the drivers of Language Technology for Catalan. Find out all about BSC’s research into the state of Spanish and Catalan in the digital world and how ELE helps "level the field for all languages in the digital arena" in the final section of this newsletter.

With best regards

Georg Rehm

• “Google is Ending Transactions for Conversational Actions in Most of Europe” – Voicebot.ai, 23 March 2022
• “Complex questions as face recognition tech joins Ukraine war” – Digital Journal, 25 March 2022
• “Artificial Intelligence, Machine Learning and Society ” – I Programmer, 28 March 2022
• “TikTok’s Black Box Obscures Its Role in Russia’s War” – Wired, 28 March 2022
• “Ukraine uses Clearview AI to identify slain Russian soldiers” – The Register, 28 March 2022
• “How to Build Your Own GPT-J Playground” – Towards Data Science , 28 March 2022
• “Language models fail to say what they mean or mean what they say” – VentureBeat, 29 March 2022
• “Quantinuum Announces Updates to Quantum NLP Toolkit λambeq” – The Register, 29 March 2022
• “Respeecher Offers Celebrities Ukrainian Speaking Voice Clones for Video Messages Supporting Ukraine” – Voicebot.ai, 29 March 2022
• “Council Post: Transforming NLP capabilities into impactful business applications” – Analytics India Magazine, 30 March 2022
• “Six Steps to Responsible AI in the Federal Government” – Brookings, 30 March 2022
• “Meta AI’s open-source system attempts to right gender bias in Wikipedia biographies” – VentureBeat, 30 March 2022
• “Artificial intelligence expands its terrain as legislation and ethics rules try to catch up” – Marketplace, 31 March 2022
• “Siri can't speak Irish: Tackling the digital gaps for the Irish language” – TheJournal.ie, 31 March 2022
• “EU has comparative advantage in autonomous robotics, AI services” – EU Science Hub, 31 March 2022
• “Europe’s AI Act contains powers to order AI models destroyed or retrained, says legal expert” – TechCrunch, 1 April 2022
• “Elevating human-machine relationships with no-code, reusable AI” – VentureBeat, 2 April 2022
• “As Russia Plots Its Next Move, an AI Listens to the Chatter” – Wired, 4 April 2022
• “Why is Meta AI’s textless NLP a breakthrough?” – Analytics India Magazine, 4 April 2022
• “Lost in translation: is research into species being missed because of a language barrier?” – The Guardian, 4 April 202

• How much AI and MT are involved in Localization? Nataly Kelly of HubSpot presents her personal estimate and clears up a few wrong conceptions about the job.
• On a similar note, ELG partner Pangeanic from Spain interviewed their Senior Project Manager about how to combine human and artificial intelligence in Language Technology and translation.
• Although the diagram is flawed, this visualisation by Lex Fridman gives an interesting perspective on the number of native speakers of major languages. The comment section is worth a visit as well.
• What do you think will be achieved sooner – Artificial General Intelligence or reliable video conferencing?

The ELG initiative continues working on a web interface for its machine translation models that translate between Ukrainian and other languages, including English, French and Russian. Last week, the team met to discuss the current design of the interface, which targets mobile devices first, but will also be available on desktop, as well as the details of the process of calling services through ELG: Via the PythonSDK, several MT models by different providers that are hosted in or are available through ELG can be selected and used for translation. The test phase for the interface starts this week and the ELG team hopes to make the service available to the public as soon as possible.

Over 20,000 responses to our ELE survey on the digital support for European languages have reached us so far – but we are aiming higher: In order to achieve an unprecedented representation of the opinion of European citizens on the support of their languages in the digital world, we have extended the deadline for the survey. Please support this initiative by making your voice heard for the sake of Digital Language Equality in Europe and sharing the survey with as many people as possible. You can choose to fill in the survey in one of 35 languages on our ELE website, while many more European regional and minority languages can be selected within the survey itself. Please participate and share the survey with your friends, colleagues and networks.

“As much as people might want to use Irish in their daily lives, the technology is almost forcing them into English. That’s just a simple example of what’s going to happen across Europe, if something doesn’t change in terms of the technologies made available.” In an article by TheJournal.ie, Dr Teresa Lynn of the ADAPT Centre at the Dublin City University, partner of the ELE project, speaks about “tackling the digital gaps for the Irish language” and the ambitious initiative for Digital Language Equality in Europe. “If you’re in the field, you know that means there’s a lot of work to be done in the next eight years.”

• June 8-9: META-FORUM 2022 (hybrid event, Brussels & online)
• June 20-25: 13th Language Resources and Evaluation Conference (LREC)

Barcelona Supercomputing Center

The Barcelona Supercomputing Center (BSC) is a public research centre located in Barcelona. It houses the MareNostrum, the most powerful supercomputer in Spain and one of the most powerful in Europe. BSC specialises in high-performance computing, aka HPC, and manages the Spanish Supercomputation Network, with the mission to offer supercomputing infrastructure and services to Spanish and European scientists, and to generate knowledge and technology in a variety of scientific fields. With this purpose, it actively participates in the main European HPC initiatives in close cooperation with other European supercomputing centres. With a total staff of more than 760 R&D experts and professionals, BSC has been successful in attracting talent. The main research lines, focused on the fields of Computer Sciences, Life Sciences, Earth Sciences and Computer Applications, are developed within the framework of the EU research funding programmes, Spanish and Catalan public research calls and collaborations with leading companies.

The Text Mining Unit at BSC has a consolidated experience in several NLP areas, such as massive language model building, biomedical text mining, machine translation and unsupervised learning for under-resourced languages and domains. It has been entrusted by the Spanish and the Catalan government with the mission to develop fundamental open-source resources and technologies for Spanish and Catalan. In connection with this, the TeMu Unit is currently in charge of two flagship projects at the national and regional level: the Spanish National Language Technology Plan, funded by the Spanish Secretariat of Digitalisation and Artificial Intelligence, and the AINA project, aimed at developing AI resources for Catalan, funded by the Catalan Digitalisation Department. In addition, the Unit participates in various EU funded international projects.

BSC in the ELE project

Within the ELE project, TeMu has been in charge of running a comprehensive survey about existing tools and resources for Spanish and Catalan, and of compiling two reports about how each of these languages are facing the digital challenge. Spanish, being one of the most spoken languages in the world, is not threatened by globalisation in the way Catalan and other languages are. However, the gap in the number of resources and tools compared to English is still quite large.

The full reports can be downloaded from the ELE website: in the Spanish and Catalan sections, respectively, and the complete registry of resources has been documented in the European Language Grid (ELG), where further details can be consulted and the resources accessed.

Maite Melero, Head of the Machine Translation Research group:

The next ELT newsletter will be sent out on 26 April 2022. Until then, follow our ELT social media accounts (as linked below) for the latest news!