Wikipedia based dataset to train relationship classifiers and fact extraction models - View it on GitHub