Learning CCG parsers for mapping natural language to (A)MDP reward functions. - View it on GitHub