欧洲人工智能的语言数据：双线作战.pdf

《欧洲人工智能的语言数据：双线作战.pdf》由会员分享，可在线阅读，更多相关《欧洲人工智能的语言数据：双线作战.pdf（10页珍藏版）》请在三个皮匠报告上搜索。

1、Christos EllinidesDirector-General Directorate-General for Translation,European CommissionLanguage Data for European AI:Striking on Two FrontsLanguage Data for European AI:Striking on Two FrontsStriking on Two FrontsStriking on Two Fronts1.Modern and technologically revamped Public Administrations2.

2、EU digital and technical sovereignty People are at the heart of translationOver 70-year historyDGT translates into and out of 24 official EU languagesCurrent output:2.5 million pages per yearIntroduction to DGTLanguage Data for European AI:Striking on Two FrontsLanguage Data for European AI:Striking

3、 on Two FrontsEU Collaboration on AI European sovereignty&wealth of languages25+Years of Data Translation archivesAI-based services and Machine Translation Evolution of data usageMultilingual Dataset Includes low-resource languagesDGTs Role in AI&Language DataDGTs translations throughout the yearsMa

4、chine translation and AI based multilingual servicesEuropean AI developmentFrom Translations to AI-based servicesLinguistic data assets:(Euramis and DGT-TM)2 billion sentences 300 000 a dayDGT-TMLaunched in 2007 used by research,business,freelance translators and moreEuramis data repositoryLaunched

5、in 1995Language Data for European AI:Striking on Two FrontsExpanding into language technologiesMachine translation MTEC-eTranslationELRC Data Collection(2016-2020)special support for low-resource languagesBroadening reach(expansion into non-translation applications:AI services)Language Data for Euro

6、pean AI:Striking on Two FrontsEuropean HPC Joint UndertakingEuropean Language Data Space European marketplace for language data supporting Europes digital sovereignty MeluXina and Leonardo supercomputers EU institutional LLMmade in Europeop