Publication:
CheckMate: English grammatical error correction for native Turkish speakers

dc.contributor.authorErsoy, Asım
dc.contributor.authorSavlak, O.
dc.contributor.authorKart, C.
dc.contributor.authorYıldız, Olcay Taner
dc.contributor.departmentComputer Science
dc.contributor.ozuauthorYILDIZ, Olcay Taner
dc.contributor.ozugradstudentErsoy, Asım
dc.date.accessioned2023-11-07T08:32:22Z
dc.date.available2023-11-07T08:32:22Z
dc.date.issued2023
dc.description.abstractThe covid-19 outbreak left many countries have no choice but turn to online education. Turkish students, who were among those who were affected, faced difficulties in improving their English as the opportunity to have face-to-face feedback was not available. In this work, we build and open-source a dataset for grammatical error correction composed of essays written by Turkish students from different universities capturing the errors Turkish native speakers tend to make. We utilize the dataset and build a model, which we deploy along with a web interface.en_US
dc.identifier.doi10.1109/SIU59756.2023.10223761en_US
dc.identifier.isbn979-835034355-7
dc.identifier.scopus2-s2.0-85173488925
dc.identifier.urihttp://hdl.handle.net/10679/8941
dc.identifier.urihttps://doi.org/10.1109/SIU59756.2023.10223761
dc.identifier.wos001062571000018
dc.language.isoengen_US
dc.publicationstatusPublisheden_US
dc.publisherIEEEen_US
dc.relation.ispartof2023 31st Signal Processing and Communications Applications Conference (SIU)
dc.relation.publicationcategoryInternational
dc.rightsrestrictedAccess
dc.subject.keywordsData collectionen_US
dc.subject.keywordsData labelingen_US
dc.subject.keywordsDeep learningen_US
dc.subject.keywordsGrammatical error correctionen_US
dc.titleCheckMate: English grammatical error correction for native Turkish speakersen_US
dc.typeconferenceObjecten_US
dspace.entity.typePublication
relation.isOrgUnitOfPublication85662e71-2a61-492a-b407-df4d38ab90d7
relation.isOrgUnitOfPublication.latestForDiscovery85662e71-2a61-492a-b407-df4d38ab90d7

Files

License bundle

Now showing 1 - 1 of 1
Placeholder
Name:
license.txt
Size:
1.45 KB
Format:
Item-specific license agreed upon to submission
Description:

Collections