EVALIGN: Visual Evaluation of Translation Alignment Models

Tariq Yousef*, Gerhard Heyer, Stefan Jänicke

*Kontaktforfatter

Publikation: Kapitel i bog/rapport/konference-proceedingKonferencebidrag i proceedingsForskningpeer review

Abstract

This paper presents EVALIGN, a visual analytics framework for quantitative and qualitative evaluation of automatic translation alignment models. EVALIGN offers various visualization views enabling developers to visualize their models’ predictions and compare the performance of their models with other baseline and state-of-the-art models. Through different search and filter functions, researchers and practitioners can also inspect the frequent alignment errors and their positions. EVALIGN hosts nine gold standard datasets and the predictions of multiple alignment models. The tool is extendable, and adding additional datasets and models is straightforward.

OriginalsprogEngelsk
TitelProceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics : System Demonstrations
Antal sider21
ForlagAssociation for Computational Linguistics (ACL)
Publikationsdato2023
Sider277-297
ISBN (Elektronisk)9781959429456
StatusUdgivet - 2023
Begivenhed17th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2023 - Dubrovnik, Kroatien
Varighed: 2. maj 20234. maj 2023

Konference

Konference17th Conference of the European Chapter of the Association for Computational Linguistics, EACL 2023
Land/OmrådeKroatien
ByDubrovnik
Periode02/05/202304/05/2023
SponsorAdobe, Babelscape, Bloomberg Engineering, Duolingo, LivePerson

Fingeraftryk

Dyk ned i forskningsemnerne om 'EVALIGN: Visual Evaluation of Translation Alignment Models'. Sammen danner de et unikt fingeraftryk.

Citationsformater