本演講將會介紹目前應用統計式的方法於語音對話系統，以及對於更開放式的領域的應用。與傳統的規則式方法不同，近年來的研究方向更著重於監督式的統計模型的訓練以及運用強化式機器學習的方法，讓語音對話系統的效能更加提升。

Abstract:

This talk will discuss methods by which current statistical approaches to spoken dialogue can be extended to cover much wider domains. In contrast to traditional rule-based approaches to building spoken dialogue systems (SDS), recent research has shown that it is possible to implement all of the required functionality using statistical models trained using a combination of supervised learning and reinforcement learning.