Article credit Slator.
To the European Union language equality is an important cause. However, many languages lack the technological support and situational context needed to continue to exist and prosper as living languages.
A strategic plan and roadmap have been developed by the European Language Equality (ELE) project to address this issue in a comprehensive way and achieve digital language equality in Europe by 2030. At a STOA Workshop held in the European Parliament in Brussels in November 2022, ELE’s Strategic Research and Innovation Agenda (SRIA) was presented.
Now, ELE is calling for support from the language community to endorse the SRIA. “This is an outreach effort to mobilize support and awareness to promote ELE’s goals that ultimately benefit all citizens,” Federico Gaspari, a research fellow at ADAPT working on the ELE project, told Slator.
The roadmap and the SRIA are the foundation for a “large-scale, long-term coordinated funding programme for research, development and innovation in the field of language technologies, at European, national and regional levels, tailored specifically to Europe’s needs and demands”, as specified by the European Parliament Resolution on language equality in the digital age in 2018.
“The SRIA is based on a careful mapping of all digital resources, tools, and technologies available for all languages in Europe from the European Language Grid (ELG) Catalogue,” explained Gaspari.
The quantification of the level of technological support for each language is enabled through the Digital Language Equality (DLE) Metric, a measure of a language’s digital readiness and its contribution to technology-enabled multilingualism. It considers: a) technological factors (such as language resources and tools and services) and b) contextual factors (such as societal, economic, educational, and industrial conditions). The DLE Metric can be computed dynamically and the scores can be interactively visualized in real time via the ELE/ELG Dashboard.
The results reveal that English is still by far the language with the best and most thorough technological support, while the majority of other European languages suffer from a lack of technological support. More specifically, more than half of the 90 European languages have either poor or no technological support at all and the gap between English and the other languages appears to be widening rather than narrowing, as mentioned in the SRIA.
This is an imbalance that the ELE project seeks to address. “We believe that all European languages should be supported, respected, and served equally,” said Gaspari. He also highlighted that another important reason for all languages being brought to a level where they are sufficiently digitally supported is to take advantage of the important progress in artificial intelligence (AI).
All languages, with the exception of English, German and French, exist in socio-political and economical ecosystems that do not incentivize, encourage or foster the development of technologies for these languages. Furthermore, almost all of the national AI strategies put in place by the 30 European countries surveyed seem to have either ignored or left out the topic of languages and language-centric AI.
The ELE project aims to reduce the technology gap between English and all other European languages and address the lack of available language data, with a focus on openness, interoperability, and standardization. The ELE project makes use of and strengthens existing as well as emerging infrastructures and data spaces.
With regard to the scientific dimension, the ELE project attempts to achieve the goal of deep natural language understanding (NLU) by 2030. Emphasis is on the development of open-access large language models (LLMs) for all European languages including the development of datasets, multilingual models, models that include symbolic knowledge, discourse features, grounding, as well as other sophisticated features that are currently beyond the capabilities of the current state-of-the-art technologies.
Finally, the ELE project will not only achieve digital language equality in Europe but also “move European research and industry in this field into a dominating position for years to come,” according to the SRIA.
Advertisement: PhD Studentship in Harmonised Semantic Model of EU AI and Data Protection Regulations. Supervised by @coolharsh55 and @EdoCeles at @AdaptCentre in @DCU. Details and application: https://www.adaptcentre.ie/careers/phd-studentship-in-harmonised-semantic-model-of-eu-ai-and-data-protection-regulations/ Please repost 🙏