DeepMind Q&A Dataset – New York University

Hermann et al. (2015) created two awesome datasets using news articles for Q&A research. Each dataset contains many documents (90k and 197k each), and each document companies on average 4 questions approximately. Each question is a sentence with one missing word/phrase which can be found from the accompanying document/context.

The original authors kindly released the scripts and accompanying documentation to generate the datasets (see here). Unfortunately due to instability of WaybackMachine, it is often cumbersome to generate the datasets from scratch using the provided scripts. Furthermore, in certain parts of the world, it turned out to be far from being straight-forward to access the WaybackMachine.

I am making the generated datasets available here. This will hopefully make the datasets used by a wider audience and lead to faster progress in Q&A research.

Hermann, K. M., Kocisky, T., Grefenstette, E., Espeholt, L., Kay, W., Suleyman, M., & Blunsom, P. (2015). Teaching machines to read and comprehend. In Advances in Neural Information Processing Systems (pp. 1684-1692).

Continue reading here:

DeepMind Q&A Dataset - New York University

Comments are closed.