European Language Resource Coordination
European Language Resource Coordination
The European Language Resource Coordination (ELRC) focuses on the collection of language data, so-called language resources, in order to improve the machine translation system of the European Commission.
Aim of ELRC
ELRC aims at identifying, collecting and processing language resources in all countries participating in the Connecting Europe Facility funding initiative, i.e. all EU member states as well as Iceland and Norway. These language resources are used to improve eTranslation, the machine translation service within the European Commission's CEF Automated Translation Platform. eTranslation aims to support cross-border, multilingual communication between public institutions, ministries and EU citizens. However, the quality of automated translation depends largely on the quality and quantity of language resources that were used to train the system. Therefore, the provision of data for a high-quality translation in the respective language is crucial.
Language resources
ELRC collects bilingual or multilingual content in digital and editable formats. These data provide the best basis for extending eTranslation's coverage and improving its performance. Ideally, data donations will include translation memories, e.g. as TMX files, or alternatively documents in the original format with source references and at least one translated document.
The following language data are particularly helpful:
- Internal reports and other documents
- Publications and other information material for external use
- Web content and brochures
- Terminology databases and glossaries
Quality is the key and quantity the way to success
The bilingual and multilingual data collected by ELRC and made available by public administrations are mainly provided by professional translators. This ensures the high quality of the content provided in the ELRC-SHARE repository. A large amount of language data contributes to the content coverage of essential topics and also promotes the consistency of the generated translation. The more frequently certain expressions are used in a specific context, the more precise the automatically translated texts become. This reduces post-editing efforts while improving the quality of machine translation.
ELRC Services
In order to support data donors who want to make their language resources available and thus contribute to the improvement of machine translation systems in CEF eTranslation, the ELRC consortium offers assistance at every step to make the data sharing process as easy as possible.
ELRC offers the following services free of charge to all data donors in CEF countries:
- Data extraction and conversion
- Removing tags and reformatting
- Data cleansing and alignment
- anonymisation
- and much more.
National Anchor Points
Each country has two National Anchor Points (NAPs): a Technological National Anchor Point (Technology NAP) and a Public Services National Anchor Point (Public Services NAP).
The Technology NAP is a renowned language or language technology expert. He or she often has an excellent academic or research background or represents a national language institution.
The Public Services NAP is a representative of the national public service, public administration or ministry. He or she acts as a point of contact for national, regional and local administrations and is able to disseminate and mobilise the importance of language resources and the performance of the ELRC among the authorities/ministries of each country.
The NAPs in Austria
- Technology NAP (at the Centre for Translation Studies): Dagmar Gromann
- Technology NAP - Substitute (Centre for Translation Studies): Barbara Heinisch
- Public Services NAP (at the Language Institute of the Federal Armed Forces): Jürgen Kotzian
Further information: http://www.lr-coordination.eu/