Сложности предметов
metric
ID: itemdifficulties
Related Papers
Comparing Bayesian Models of Annotation
AbstractThe analysis of crowdsourced annotations in natural language processing is concerned with identifying (1) gold standard labels, (2) annotator accuracies and biases, and (3) item difficulties and error patterns. Traditionally, …...