“TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension”, Mandar Joshi, Eunsol Choi, Daniel S. Weld, Luke Zettlemoyer2017-05-09 (ML dataset, RNN)⁠:

We present TriviaQA, a challenging reading comprehension dataset containing over 650K question-answer-evidence triples. TriviaQA includes 95K question-answer pairs authored by trivia enthusiasts and independently gathered evidence documents, 6 per question on average, that provide high quality distant supervision for answering the questions.

We show that, in comparison to other recently introduced large-scale datasets, TriviaQA (1) has relatively complex, compositional questions, (2) has considerable syntactic and lexical variability between questions and corresponding answer-evidence sentences, and (3) requires more cross sentence reasoning to find answers.

We also present two baseline algorithms: a feature-based classifier and a state-of-the-art neural network, that performs well on SQuAD reading comprehension. Neither approach comes close to human performance (23% and 40% vs. 80%), suggesting that TriviaQA is a challenging testbed that is worth future study.

Data and code available at https://nlp.cs.washington.edu/triviaqa/.