NewsQA: A Machine Comprehension Dataset

TLDR

NewsQA is a challenging machine comprehension dataset comprising over 100,000 human‑generated question‑answer pairs. The dataset was built by crowdworkers who generated questions and span‑based answers from more than 10,000 CNN news articles through a four‑stage process that encourages exploratory, reasoning‑requiring questions, and human performance was benchmarked against strong neural models. Analysis shows NewsQA requires reasoning beyond word matching, with a 13.3 % F1 gap between humans and state‑of‑the‑art neural models, and the dataset is freely available online.

Abstract

We present NewsQA, a challenging machine comprehension dataset of over 100,000 human-generated question-answer pairs. Crowdworkers supply questions and answers based on a set of over 10,000 news articles from CNN, with answers consisting of spans of text in the articles. We collect this dataset through a four-stage process designed to solicit exploratory questions that require reasoning. Analysis confirms that NewsQA demands abilities beyond simple word matching and recognizing textual entailment. We measure human performance on the dataset and compare it to several strong neural models. The performance gap between humans and machines (13.3% F1) indicates that significant progress can be made on NewsQA through future research. The dataset is freely available online.