Are We Modeling the Task or the Annotator? An Investigation of Annotator Bias in Natural Language Understanding Datasets.

Mor Geva, Yoav Goldberg, Jonathan Berant

Browse the full EMNLP paper archive.