Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

Paragraph-Level Translation of Low-Resource Finno-Ugric Languages

Authors: Pashchenko, Dmytro;

Paragraph-Level Translation of Low-Resource Finno-Ugric Languages

Abstract

Massiliselt mitmekeelsete masintõlkemudelite teke võimaldas tõhusalt tõlkida paljusid keeli samaaegselt, sealhulgas neid, millel on piiratud hulk ressursse. Hiljutine rekordiomanik MADLAD-400, mis katab üle 400 keele, on suuresti uurimata. Käesolevas töös püüame uurida MADLADi võimekust, häälestades seda nelja väikese ressursiga soomeugri keele (karjala, liivi, lüüdi ja vepsa, mis ei sisaldu MADLADis) tõlkimisele vene keelde ja tagasi. Lisaks uurime lõigutasandil tõlke mõju mudeli kvaliteedile, kasutades MADLADi dokumenditasemel tõlkimise võimekust. Leiame, et 1) MADLADi-põhine süsteem saavutab tipptasemel mudelitega võrreldavad tulemused ja avastame, et 2) süsteemi lõigutasandil versioon ületab lausetasemel versiooni kuni kolme BLEU punkti võrra, parandades oluliselt lausetevahelist kooskõla.

Countries
Germany, Estonia
Related Organizations
Keywords

massively multilingual models, paragraph-level translation, MADLAD-400, discourse-level phenomena, informatics, magistritööd, infotehnoloogia, infotechnology, neural machine translation, informaatika, 004

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green