Pregled bibliografske jedinice broj: 827066
Multilingual projection for parsing truly low-resource languages
Multilingual projection for parsing truly low-resource languages // Transactions of the Association for Computational Linguistics, 4 (2016), 301-312 (podatak o recenziji nije dostupan, članak, znanstveni)
CROSBI ID: 827066 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Multilingual projection for parsing truly low-resource languages
Autori
Agić, Željko ; Johannsen, Anders ; Plank, Barbara ; Alonso Martínez, Héctor ; Schluter, Natalie ; Søgaard, Anders
Izvornik
Transactions of the Association for Computational Linguistics (2307-387X) 4
(2016);
301-312
Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni
Ključne riječi
dependency parsing ; cross-lingual ; low-resource languages
Sažetak
We propose a novel approach to cross-lingual part-of- speech tagging and dependency parsing for truly low- resource languages. Our annotation projection-based approach yields tagging and parsing models for over 100 languages. All that is needed are freely available parallel texts, and taggers and parsers for resource-rich languages. The empirical evaluation across 30 test languages shows that our method consistently provides top-level accuracies, close to established upper bounds, and outperforms several competitive baselines.
Izvorni jezik
Engleski
Znanstvena područja
Računarstvo
