On International Mother Language Day: Upper Sorbian in the Microsoft Bing translator
21.02.2022
Domowina, press release from February 21, 2022:
This year's mother tongue day, which is celebrated internationally on February 21, is a very special one for the Upper Sorbian language spoken in the Saxon region of Upper Lusatia: it is the first minority language in Germany to be integrated into Microsoft's Bing translator. There, Upper Sorbian can not only be translated into German, but also into over a hundred other languages - and vice versa.
Individual sentences or entire documents can also be translated in other Microsoft applications such as Office 365. It is based on AI technology that learns from the language data. The more the system is used, the better the results will be.
As early as 2014, the WITAJ language center of the Domowina received the suggestion from Microsoft employees as part of a training event to develop a bilingual corpus with parallel sentences in Sorbian and German. The message that such parallel data will be the basis for many digital applications in the future was well received.
The work on the parallel corpus aroused the particular interest of an IT specialist, himself a Sorbian learner. Together with the WITAJ language center, he began to use the small corpus for translation experiments ̶ with success! The Foundation for the Sorbian People has been funding the development of the translation program since 2019. In March 2021, the first version of "sotra" was published - the first online translation program Upper Sorbian - German and German - Upper Sorbian. Scientifically sound recommendations for further development came from Prof. Dr. Alexander Fraser from the Center for Information and Language Processing at the Ludwig Maximilian University of Munich.
Microsoft would like to promote the preservation of languages worthy of protection. The fact that Upper Sorbian has now also been included in the Bing translator and has thus been integrated for the first time by a large corporation into one of its translation services is the result of solid and persistent teamwork and also the commitment of the Sorbian language community.
And how does Upper Sorbian sound? Read the same message in Upper Sorbian for yourself.
K dnjej maćeršćiny: Hornjoserbšćina w Microsoft Bing-přełožowarju
Lětuši dźeń maćeršćiny, kotryž so mjezynarodnje 21. februara swjeći, je cyle wosebity za hornjoserbsku rěč: Jako prěnja mjeńšinowa rěč w Němskej so wona do Bing-přełožowarja koncerna Microsoft integruje. Tam njemóža so serbske teksty potom jenož do němčiny, ale tež do wjace hač sto druhich rěčow přełožować – a nawopak. Přełožować hodźa so jednotliwe sady abo cyłe dokumenty, tež w druhich nałožowanjach Microsofta kaž Office365. Zakład twori KI-technologija („kumštna inteligenca“), kotraž wuknje z rěčnych datow. Čim bóle so system wužiwa, ćim lěpše budu wuslědki.
Hižo w lěće 2014 bě Rěčny centrum WITAJ Domowiny we wobłuku dalekubłanskeho zarjadowanja wot sobudźěłaćerjow Microsoft nastork dóstał, dwurěčny korpus z paralelnymi sadami w serbšćinje a němčinje nadźěłać. Poselstwo, zo budu tajke paralelne daty přichodnje zakład mnohich digitalnych nałožowanjow, je RCW přiwzał.
Dźěło na paralelnym korpusu zbudźi wosebity zajim IT-specialista, kotryž sam serbsce wuknje. Zhromadnje z RCW započa wón, mały korpus za přełožowanske eksperimenty wužiwać – z wuspěchom! Wot 2019 spěchuje Załožba za serbski lud wuwiwanje přełožowanskeho programa. W měrcu 2021 wozjewi so prěnja wersija „sotry“ – prěnjeho onlinepřełožowarja za hornjoserbšćinu a němčinu. Wědomostnje fundowane doporučenja k dalewuwiću přińdu wot prof. dr. Alexandera Frasera z Centruma za předźěłanje informacijow a rěče na Ludwiga Maximilianowej uniwersiće w Mnichowje.
Microsoft chce wobchowanje małych rěčow spěchować. Zo bu hornjoserbšćina nětko tež do Bing-translatora přiwzata a z tym prěni raz wot jednoho z wulkich koncernow do přełožowanskeho poskitka integrowana, je wuslědk solidneho a wutrajneho teamoweho dźěła a tež angažementa Serbow samych.
