Fully Attentional Network for Low-Resource Academic Machine Translation and Post Editing

Basit Öğe Kaydı
dc.authoridsel, ilhami/0000-0003-0222-7017
dc.authoridHanbay, Davut/0000-0003-2271-7865
dc.authorwosidsel, ilhami/ABD-7350-2020
dc.authorwosidHanbay, Davut/AAG-8511-2019
dc.contributor.authorSel, Ilhami
dc.contributor.authorHanbay, Davut
dc.date.accessioned2024-08-04T20:53:12Z
dc.date.available2024-08-04T20:53:12Z
dc.date.issued2022
dc.departmentİnönü Üniversitesien_US
dc.description.abstractEnglish is accepted as an academic language in the world. This necessitates the use of English in their academic studies for speakers of other languages. Even when these researchers are competent in the use of the English language, some mistakes may occur while writing an academic article. To solve this problem, academicians tend to use automatic translation programs or get assistance from people with an advanced level of English. This study offers an expert system to enable assistance to the researchers throughout their academic article writing process. In this study, Turkish which is considered among low-resource languages is used as the source language. The proposed model combines the transformer encoder-decoder architecture model with the pre-trained Sci-BERT language model via the shallow fusion method. The model uses a Fully Attentional Network Layer instead of a Feed-Forward Network Layer in the known shallow fusion method. In this way, a higher success rate could be achieved by increasing the attention at the word level. Different metrics were used to evaluate the created model. The model created as a result of the experiments reached 45.1 BLEU and 73.2 METEOR scores. In addition, the proposed model achieved 20.12 and 20.56 scores, respectively, with the zero-shot translation method in the World Machine Translation (2017-2018) test datasets. The proposed method could inspire other low-resource languages to include the language model in the translation system. In this study, a corpus composed entirely of academic sentences is also introduced to be used in the translation system. The corpus consists of 1.2 million parallel sentences. The proposed model and corpus are made available to researchers on our GitHub page.en_US
dc.description.sponsorshipInonu University scientific research and coordination unit [FDK-2022-2925]en_US
dc.description.sponsorshipThis study was funded by the Inonu University scientific research and coordination unit with the Project number FDK-2022-2925.en_US
dc.identifier.doi10.3390/app122211456
dc.identifier.issn2076-3417
dc.identifier.issue22en_US
dc.identifier.scopus2-s2.0-85142528471en_US
dc.identifier.scopusqualityQ2en_US
dc.identifier.urihttps://doi.org/10.3390/app122211456
dc.identifier.urihttps://hdl.handle.net/11616/101013
dc.identifier.volume12en_US
dc.identifier.wosWOS:000887162200001en_US
dc.identifier.wosqualityQ2en_US
dc.indekslendigikaynakWeb of Scienceen_US
dc.indekslendigikaynakScopusen_US
dc.language.isoenen_US
dc.publisherMdpien_US
dc.relation.ispartofApplied Sciences-Baselen_US
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Kurum Öğretim Elemanıen_US
dc.rightsinfo:eu-repo/semantics/openAccessen_US
dc.subjectnatural language processingen_US
dc.subjectneural machine translationen_US
dc.subjecttransformeren_US
dc.subjectfully attentional networken_US
dc.subjectparallel corpusen_US
dc.titleFully Attentional Network for Low-Resource Academic Machine Translation and Post Editingen_US
dc.typeArticleen_US

Dosyalar

Koleksiyon

WoS İndeksli Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu