The technical objective of EUROTRA is the creation of a prototype machine translation system capable of dealing with all nine official Community languages. Linguistic models are to be developed for the analysis and synthesis of each language and for transfer between them, initially using a corpus with a vocabulary of about 2500 entries from a limited specialised field. A lexical database must be prepared for this vocabulary. Linguistic strategies will be developed for machine- execution of various processes and a "rule interpreter" developed as part of the basic software. The translation process is formalised as a succession of textual mappings through a number of linguistically motivated grammatical representations. The mapping from one level of representation to the next has to be describable by a set of rules (translators). There is one analysis and one synthesis package for each language. The project has a decentralised structure, with national EUROTRA centres.
Bibliographic Reference: EUR 12853 EN (1990) 12 PP., FS, FREE OF CHARGE ISBN 92-826-1441-7 THE OFFICE FOR OFFICIAL PUBLICATIONS OF THE EUROPEAN COMMUNITIES, L-2985 LUXEMBOURG
Availability: Can be ordered online
Record Number: 1989128070400 / Last updated on: 1990-11-09
Available languages: en,de,es,fr