Learning non-cooperative dialogue behaviours

摘要: Non-cooperative dialogue behaviour has been identified as important in a variety of application areas, including education, military operations, video games and healthcare. However, it not addressed using statistical approaches to management, which have always trained for co-operative dialogue. We develop evaluate agent learns perform noncooperative moves order complete its own objectives stochastic trading game. show that, when given the ability both cooperative non-cooperative moves, such an can learn bluff lie so win more often ‐ against adversaries, under various conditions risking penalties being caught deception. For example, we that additional 15.47% strong rulebased adversary, compared optimised cannot moves. This work is first how use effectively meet goals.