: Vocabulario de PLN: score

NLP resources and evaluation > measure > metric > score

score

Unlike majority of the previous work we create a framework for training large number of features on human rankings and at the same time as a result of tuning produce a score based metric which does not require two (or more) hypotheses for comparison. (Stanojević, Miloš and Sima’an, Khalil, BEER: BEtter Evaluation as Ranking, Proceedings of the Ninth Workshop on Statistical Machine Translation, Association for Computational Linguistics (https://www.aclweb.org/anthology/W14-3354), 2014)

http://data.loterre.fr/ark:/67375/8LP-M5GJKWH3-1

RDF/XML TURTLE JSON-LD última modificación 2/5/24

Loterre