Poul Andersen * Cooperation with Central and Eastern Europe in Language Engineering

7. Joint Research Projects

These projects aim to assemble, for a specific research subject, a multinational team to perform research and development work and to obtain results in collaboration. A Joint Research Project typically comprises 3-6 partners coming from at least 3 different countries in Eastern and Western Europe. The duration of a project is normally between one and three years.

In the area of Language Engineering, 10 Joint Research Projects involving Central & Eastern Europe started in 1995. Four of these projects have Romanian participation:

• ONOMASTICA-COPERNICUS

  The project ONOMASTICA is building a pronunciation lexicon for the names of the European Union. These are city and town names, street names, family names, product names in a total of 11 languages - Danish, Dutch, English, French, German, Greek, Italian, Norwegian, Portuguese, Spanish and Swedish.

  The goal for ONOMASTICA-Copernicus is to extend the languages covered in the database to include names and pronunciations for Czech, Estonian, Latvian, Polish, Romanian, Slovakian, Slovenian and Ukrainian.

  Pronunciation dictionaries for up to 250,000 names per language will be constructed. The objective of the project is to make available quality controlled pronunciation lexicons in machine readable form (CD-ROM) for use in automated language systems, of primary interest to international European companies in the telecommunications sector and in the (dictionary) publishing industry, as well as to language system researchers and developers.

  The Romanian partner in ONOMASTICA-Copernicus is Prof. Cezar Tabarcea, Faculty of Arts, University of Bucharest.
  
  

• MULTEXT-EAST

  MULTEXT-EAST is a spin-off of MULTEXT, one of the largest EU projects in the domain of language tools and resources. MULTEXT has three main objectives:
  
  

  • Standardization: development of a software standard based on a "software Lego" approach for corpus handling tools, together with TEI-based encoding conventions specifically suited to multilingual corpora and language engineering applications.
  • Tool and corpus development: development of an extensive set of tools for corpus annotation and exploitation as well as the first annotated large-scale multilingual corpus for EU languages, intended to serve as a reference and test-bed for multilingual tools and applications.
  • Industrial validation: integration by six major European companies of project results into high-level NLP applications such as term extraction and machine translation lexicon generation, thus providing a first indication of downstream applicability.

  MULTEXT-EAST extends the scope of MULTEXT to CEE countries, and together the two projects create a network of more than 20 academic research centres and companies, developing and using common lingware and methodologies, as well as producing the first annotated large-scale multilingual corpus for 12 EU and CEE languages.

  East European languages covered: Bulgarian, Czech, Estonian, Hungarian, Romanian, Slovenian.

  The Romanian partner in MULTEXT-EAST is the same as in TELRI (see above).
  
  

• PRACTEAST - Preparatory Actions for Terminological Assistance to Central and Eastern European Countries

  The concrete and practical outcome of PRACTEAST is the compilation of four terminological collections in such priority domains as Economics and Management, Energy, Environment and Telecommunications, each collection containing the 2,000 most common terms in the field with English terms and definitions and French and Spanish equivalent terms.

  Each partner will be able to use the Multilingual Database for its own research purposes. The corresponding conventional paper dictionaries can be published for use by the public, and each associated partner can benefit from the possible sales of such dictionaries.

  East European languages covered: Bulgarian, Czech, Estonian, Hungarian, Latvian, Lithuanian, Polish, Romanian, Russian, Slovakian and Ukrainian.

  The Romanian partners are Marius Sala, Institute of Linguistics (Romanian Academy of Sciences), and Mihail Ciocodeica, Romanian Standards Institute.

16