NorQuAD is the first Norwegian question answering dataset specifically designed for machine reading comprehension. This dataset comprises 4,752 manually created question-answer pairs. The data collection procedure is detailed in the paper, along with relevant statistics. Additionally, several multilingual and Norwegian monolingual language models were benchmarked against human performance using this dataset¹²³.

(1) [2305.01957] NorQuAD: Norwegian Question Answering Dataset - arXiv.org. https://arxiv.org/abs/2305.01957. (2) NorQuAD: Norwegian Question Answering Dataset - ACL Anthology. https://aclanthology.org/2023.nodalida-1.17/. (3) NorQuAD: Norwegian Question Answering Dataset. https://researchportal.helsinki.fi/en/publications/norquad-norwegian-question-answering-dataset. (4) undefined. https://doi.org/10.48550/arXiv.2305.01957.