Free917
As one of the earliest attempts to build a dataset for KBQA, Free917[1] contains 917 pairs of question and formal query in Freebase. In this dataset, 81 domains are represented, and the lambda calculus forms contains 635 distinct Freebase relations. The most common domains, film and business, each took up no more than 6% of the overall dataset.
Leaderboard
References
[1] Cai, Qingqing, and Alexander Yates. Large-scale semantic parsing via schema matching and lexicon extension. In Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 423-433. 2013.