Question Answer Dataset V1.2

Question Answer Dataset is a question-answering dataset that combines links to a corpus of Wikipedia articles and manually generates simulated questions and answers for academic research. The dataset is divided into three parts: articles, questions, and answers. The manually generated simulated questions and answers are very difficult to score on Wikipedia articles.

Question Answer Dataset was released by Carnegie Mellon University in 2013. The main publishers were Noah Smith, Michael Heilman, Rebecca Hwa, Shay Cohen and Kevin Gimpel. The earliest version of the data was mostly collected by students from Carnegie Mellon University and University of Pittsburgh from 2008 to 2010.