Representing the Reinforcement Learning state in a negotiation dialogue | IEEE Conference Publication | IEEE Xplore