Results of the Translation Inference Across Dictionaries 2021 Shared Task

Gracia, Jorge (Universidad de Zaragoza) ; Kabashi, Besim ; Kernerman, IIan

Resumen: The objective of the Translation Inference Across Dictionaries (TIAD) shared task is to explore and compare methods and techniques that infer translations indirectly between language pairs, based on other bilingual/multilingual lexicographic resources. In this forth edition the participating systems were asked to generate new translations automatically among three languages - English, French, Portuguese - based on known indirect translations contained in the Apertium RDF graph. Such evaluation pairs have been the same during the three last TIAD editions. The main novelty this time has been the use of a larger graph as a basis to produce the translations, which is the Apertium RDF v2, and the introduction of improved evaluation metrics. The evaluation of the results was carried out by the organisers against manually compiled language pairs of K Dictionaries. For the first time in the TIAD series, some systems beat the proposed baselines. This paper gives an overall description of the shard task, the evaluation data and methodology, and the systems’ results.

Idioma: Inglés

Año: 2021

Publicado en: CEUR Workshop Proceedings 3064 (2021), 208-220
ISSN: 1613-0073

Originalmente disponible en: Texto completo de la revista

Financiación: info:eu-repo/grantAgreement/ES/AEI/PID2020-113903RB-I00
Financiación: info:eu-repo/grantAgreement/EC/H2020/731015/EU/European Lexicographic Infrastructure/ELEXIS
Financiación: info:eu-repo/grantAgreement/EC/H2020/825182/EU/Ready-to-use Multilingual Linked Language Data for Knowledge Services across Sectors/Pret-a-LLOD
Financiación: info:eu-repo/grantAgreement/ES/MINECO/RYC2019-028112-I

Tipo y forma: Artículo (Versión definitiva)

Área (Departamento): Área Lenguajes y Sistemas Inf. (Dpto. Informát.Ingenie.Sistms.)