(2016. 6) Squad

• Submitted on 2016. 6

• Pranav Rajpurkar, Jian Zhang, Konstantin Lopyrev and Percy Liang

Simple Summary

Stanford Question Answering Dataset (SQuAD), a new reading comprehension dataset consisting of 100,000+ questions posed by crowdworkers on a set of Wikipedia articles, where the answer to each question is a segment of text from the corresponding reading passage.

• https://rajpurkar.github.io/SQuAD-explorer/

• a large reading comprehension dataset on Wikipedia articles with crowdsourced.

• 107,785 question-answer pairs on 536 articles ~500 articles from Wikipedia and size 100K.

• human performance: 86.8 F1, 77% exact match

• answer types

images