15th Annual Conference of the International Speech Communication Association

September 14-18, 2014

Inverse Reinforcement Learning for Micro-Turn Management

Dongho Kim, Catherine Breslin, Pirros Tsiakoulis, M. Gašić, Matthew Henderson, Steve Young

University of Cambridge, UK

Existing spoken dialogue systems are typically not designed to provide natural interaction since they impose a strict turn-taking regime in which a dialogue consists of interleaved system and user turns. To allow more responsive and natural interaction, this paper describes a system in which turn-taking decisions are taken at a more fine-grained micro-turn level. A decision-theoretic approach is then applied to optimise turn-taking control. Inverse reinforcement learning is used to capture the complex but natural behaviours from human-human dialogues and optimise interaction without specifying a reward function manually. Using a corpus of human-human interaction, experiments show that IRL is able to learn an effective reward function which outperforms a comparable handcrafted policy.

Full Paper

Bibliographic reference.  Kim, Dongho / Breslin, Catherine / Tsiakoulis, Pirros / Gašić, M. / Henderson, Matthew / Young, Steve (2014): "Inverse reinforcement learning for micro-turn management", In INTERSPEECH-2014, 328-332.