Dear reader,

An eventful week lies behind us: after three days of discussion at the “Innovation, Technologies and Plurilingualism” online forum organised by the French Ministry of Culture under the umbrella of the French Presidency of the Council of the EU, last week concluded with a busy Friday, which saw a national workshop in Bulgaria for the European Language Grid, where almost 100 language technology enthusiasts learned all about the platform and its advantages, and a General Assembly of the European Language Equality project and its 52 consortium members.

With 47 deliverables coming up and a first draft of the strategic research, innovation and implementation agenda and roadmap to be finished soon, the ELE project consortium is as busy as ever. However, the General Assembly last Friday was devoted to mainly one thing: The Digital Language Equality Metric and its technical and contextual factors. Read on below to learn more about it.

“We are focusing an important part of our research and innovation activities on the extension of our native multi-language capabilities to further break down language barriers and provide a concrete contribution to full digital equality across Europe”, explains José Manuel Gómez-Pérez, Language Technology Research Director at expert.ai, in the company’s profile in this week’s newsletter. Scroll on to learn more about the services provided and goals set by the multinational partner of the European Language Equality project. 

With best regards

Georg Rehm

• “How AI accelerates insurance claims processing” – VentureBeat, 2 February 2022
• “Knowledge Distillation, a Methodology for Compressing Neural Networks” – Medium, 2 February 2022
• “DeepMind Claims New AlphaCode AI Programmer Outperforms GPT-3 and is Competitive With Humans” – Voicebot.ai, 3 February 2022
• "Can an AI Write a Speech Better Than a Human?" – Built In, 3 February 2022
• “Report: 91% of top data execs agree that managing unstructured language data must be addressed” – VentureBeat, 4 February 2022
• “Inside the EU's rocky path to regulate artificial intelligence” – International Association of Privacy Professionals, 9 February 2022
• “AI Incident Database Spotlights Worst Machine Translation Fails” – Slator, 9 February 2022
• “Can AI be an effective tool against disinformation and hate speech? A geopolitical view” – Diginomica, 10 February 2022
• “AI engine exposes online fraud” – VentureBeat, 10 February 2022
• "Europe needs to hasten AI in manufacturing adoption" – TechHQ, 14 February 2022

• OpenAI Chief Scientist Ilya Sutskever claims that it may be that today's large neural networks are slightly conscious – Murray Shanahan agrees: In the same sense that it may be that a large field of wheat is slightly pasta.
• February 7 was the Love Your Robot Day in the USA. Searching for the hashtag on LinkedIn results in a nice collection of the latest developments in robotics – paired with lovely messages for our mechanical friends.
• A more important date was February 11: The International Day of Women in Science. Amongst many great posts, European translation company Pangeanic presented five important women in the history of language science and technology, such as Grace Hooper.
• The differences between British and US-American English have always been a basis for debate, but also fun and laughter – unless it affects something as important as the European Crisps Act.

“We have basic processing tools like taggers, tokenizers and parsers, but also fully fledged applications for machine translations, speech synthesis, sentiment analysis – anything you can think of!” Last Friday, on the International Day of Women in Science, Penny Labropoulou (ILSP/RC Athena) presented the catalogue and functionality of the European Language Grid at the first Bulgarian dissemination event of the project, which was opened by Svetla Koeva of the Institute for Bulgarian Language, followed by an introduction to the project by Katrin Marheinecke (DFKI).

Identrics creates bespoke automation and artificial intelligence solutions, delegating organisations’ menial data and text processing tasks to machines so that humans can focus on high-end, insight-oriented work. Its solutions range from data classification and indexing to content enrichment – such as named entity recognition, sentiment analysis, all the way to automated content recommendation and generation. Identrics is a team of data science & semantic technology experts, specialised in the development of AI-based solutions for publishing, media and risk intelligence - from ready-to-use models, to fully tailored data science as a service.

Shortly before the bulk of deliverables produced in the European Language Equality project is due – 47 in total will be sent to the EU Commission by the end of February – the consortium met for the 4th General Assembly on Friday, 11 February. The main topic of the meeting with 74 participants from all over Europe was the DLE metric, whose current status of development was presented by Federico Gaspari (ADAPT Centre, DCU) and Annika Grützner-Zahn (DFKI). The lively Q&A session and discussion amongst the consortium members will contribute to an informative metric, which after its finalisation will depict an insightful ranking of the state of digital language equality in Europe – publicly accessible on the European Language Grid.

Parliamentary data is an important source of scholarly and socially relevant content – for scientists, investigative journalists as well as citizens interested in political trends and insights. Accessible, comprehensive and well-annotated parliamentary corpora are therefore crucial for the information society and the digital humanities alike. Our colleagues at CLARIN, the Common Languages and Resource Infrastructure, will host a workshop on creating, enriching and using parliamentary corpora at this year’s Language Resources and Evaluation Conference (LREC) from 20 to 25 June 2022. The third ‘ParlaCLARIN’ workshop will take place on the first conference day of the 13th LREC at Palais du Pharo in Marseille, France.

expert.ai

expert.ai (EXAI:IM) is a leading company in AI-based natural language software. With offices and R&D labs in Europe and North America, expert.ai is one of the largest and most relevant European providers of natural language understanding and natural language processing. Thanks to the accuracy of its multi-language content understanding, which can exceed 90% accuracy within completely open or domain-independent contexts, expert.ai solutions have been deployed extensively and successfully in many industries.

“Multilingual content understanding is a top priority in many organisations’ agenda”, said José Manuel Gómez-Pérez, Language Technology Research Director at expert.ai. “We are focusing an important part of our research and innovation activities on the extension of our native multi-language capabilities to further break down language barriers and provide a concrete contribution to full digital equality across Europe”.

Expert.ai provides a scalable natural language platform that is purpose-built for the unique complexity of unstructured language data. It pairs simple and powerful tools with a proven hybrid AI approach that combines symbolic and machine learning to solve real-world problems and enhance business operations at speed and scale. Organisations in insurance, banking and finance, publishing, media, science and defence rely on expert.ai to turn language into data, analyse and understand complex documents, accelerate intelligent process automation and improve decision making.

The next ELT newsletter will be sent out on 1 March 2022. Until then, follow our ELT social media accounts (as linked below) for the latest news!