SQuAD: 100,000+ Questions for Machine Comprehension of Text

TLDR

The paper introduces SQuAD, a reading comprehension dataset of over 100,000 crowdworker‑generated questions on Wikipedia passages with answers as text spans. The authors construct the dataset and analyze it using dependency and constituency tree structures to characterize the reasoning needed for answering. A logistic regression baseline achieves 51.0 % F1, far below human 86.8 %, showing the dataset is challenging, and it is freely available online.

Abstract

We present the 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. We analyze the dataset to understand the types of reasoning required to answer the questions, leaning heavily on dependency and constituency trees. We build a strong logistic regression model, which achieves an F1 score of 51.0%, a significant improvement over a simple baseline (20%). However, human performance (86.8%) is much higher, indicating that the dataset presents a good challenge problem for future research. The dataset is freely available at https://stanford-qa.com