Written by Zsolt G. Pataki with Amr Dawood,
While multilingualism is one of Europe’s greatest assets, it is also one of the most substantial challenges for the creation of a truly integrated EU. Around 80 languages are used in the EU, and language barriers have a profound effect, not only on the establishment of a common European identity, but also on worker mobility and cross-border e-commerce, trade and public services. This study, published by STOA in March 2017, charts the emergence of new technological advancements, based on increased computational power and access to big data, through which human language technologies (HLT) are becoming real solution to overcoming language barriers. However, most of the technical progress so far has been with only one language, English.
Market fragmentation and uncoordinated funding strategies, including research and industrial policies, are hindering the European HLT community. This study examines a variety of policy options for improving the situation.
Read this study on ‘Language equality in the digital age – Towards a Human Language Project’ here.
So what are HLTs?
Human language technologies (HLTs) are software systems designed to handle human language in all its forms: spoken, written or signed. HLTs are found behind many everyday digital products, since most of them use language to some extent. Although existing HLT systems are far from achieving human abilities, there are numerous possible advancements in HLTs. the technologies are even paving new ground and providing a real solution to overcoming language barriers using new developments, such as deep learning neural networks and data analytics.
The authors of this study acknowledge that technology may be a valid response to the multilingual challenge, and, inter alia, they suggest establishing the European Human Language Project as an instrument to boost European cooperation in the next generation of HLT development and to address the issue of Europe’s multilingualism. The Human Language Project, launched by TAUS in 2012, was inspired by the Human Genome Project and is an open platform of language resources and tools, involving multiple stakeholders at the European, national and regional levels in a coordinated and joint effort to move Europe into the leading position in this field.
It is believed that the Human Language Project could help contribute to the equality of European citizens in their everyday digital experience regardless of their language. The aim is to help remove language barriers and promote an even more united EU. Implementing these policies could bring vast improvements in fostering the creation and growth of competitive European firms, boosting research and industry integration, and much more.
The study also recognises that there is no single policy that can help the EU break the existing language barriers. The authors therefore propose a combination of 11 different policy options, which have been researched and assessed, and can be grouped into:
- Institutional policies, to strengthen the role of HLT within the institutional framework of multilingualism-related bodies, and to create tools to properly evaluate HLT policies.
- Research policies, to refocus and strengthen research in LT through the Human Language Project, to promote the European LT Platform for data and services, and to bridge the technology gap between European languages.
- Industry policies to foster and support the development of investment instruments and accelerator programmes targeting HLT start-ups, and to increase the availability of qualified personnel in HLT.
- Market policies to raise awareness of the benefits for companies, public bodies and citizens of the availability of online services, contents and products in multiple languages, and promote the automated translation of e-commerce websites of European SMEs.
- Public service policies through which public procurement of innovative technology and pre-commercial public procurement could encourage the translation of national and regional public websites and documents to other EU languages using HLT.
This study was widely discussed by experts and MEPs in a STOA workshop on 10 January 2017. It was then presented to a joint ITRE/CULT Committee meeting on 25 January 2017, and to the STOA Panel on 16 February 2017. As a follow-up to its publication, the Co-Chairs of the EP Minority Intergroup have invited STOA to present the study in their meeting of 15 June 2017.
To keep up to date with this project and other STOA activities, follow our website, the EPRS blog, Twitter, and Think Tank website. We would be grateful if you could fill in the feedback questionnaire on STOA website (on the list of studies first select the title of the study and then the link ‘Your opinion counts for us’).