Natural Language Processing, Reinforcement Learning, Tempora Information