Deep Multi-Lingual Cross Sentence Alignment
Identifikátory výsledku
Kód výsledku v IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F18%3A10390127" target="_blank" >RIV/00216208:11320/18:10390127 - isvavai.cz</a>
Výsledek na webu
—
DOI - Digital Object Identifier
—
Alternativní jazyky
Jazyk výsledku
angličtina
Název v původním jazyce
Deep Multi-Lingual Cross Sentence Alignment
Popis výsledku v původním jazyce
Sentence-aligned parallel bilingual corpora are the main and sometimes the only required resource for training Statistical and Neural Machine Translation systems (SMT, NMT). We propose an end-to-end deep neural architecture for language independent sentence alignment. In addition to one-to-one alignment, our aligner can perform cross- and many-to-many alignment as well. We also present a case study which shows how simple linguistic analysis can improve the performance of a pure neural network significantly. We used three language pairs from Europarl corpus (Koehn, 2005) and an English-Persian corpus (Pilevar et al., 2011) to generate an alignment dataset. Using this dataset, we tested our system individually and in an SMT system. In both settings, we obtained significantly better results compared to an open source baseline.
Název v anglickém jazyce
Deep Multi-Lingual Cross Sentence Alignment
Popis výsledku anglicky
Sentence-aligned parallel bilingual corpora are the main and sometimes the only required resource for training Statistical and Neural Machine Translation systems (SMT, NMT). We propose an end-to-end deep neural architecture for language independent sentence alignment. In addition to one-to-one alignment, our aligner can perform cross- and many-to-many alignment as well. We also present a case study which shows how simple linguistic analysis can improve the performance of a pure neural network significantly. We used three language pairs from Europarl corpus (Koehn, 2005) and an English-Persian corpus (Pilevar et al., 2011) to generate an alignment dataset. Using this dataset, we tested our system individually and in an SMT system. In both settings, we obtained significantly better results compared to an open source baseline.
Klasifikace
Druh
O - Ostatní výsledky
CEP obor
—
OECD FORD obor
10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Návaznosti výsledku
Projekt
—
Návaznosti
S - Specificky vyzkum na vysokych skolach
Ostatní
Rok uplatnění
2018
Kód důvěrnosti údajů
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů