Reinforcement Learning (RL) algorithms provide a type of unsupervised learning that is especially well suited for the challenges of spoken dialogue systems (SDS) design. SDS are constantly subjected to new environments in the form of new groups of users, and RL provides an approach for automated learning that can adapt to new environments without costly supervision. In this paper, we describe s...

This paper addresses the problem of introducing learning capabilities in industrial handcrafted automata-based Spoken Dialogue Systems, in order to help the developer to cope with his dialogue strategies design tasks. While classical reinforcement learning algorithms position their learning at the dialogue move level, the fundamental idea behind our approach is to learn at a finer internal deci...

Adapting Spoken Dialogue Systems to the user is supposed to result in more efficient and successful dialogues. In this work, we present an evaluation of a quality-adaptive strategy with a user simulator adapting the dialogue initiative dynamically during the ongoing interaction and show that it outperforms conventional non-adaptive strategies and a random strategy. Furthermore, we indicate a co...

We describe an approach to dealing with interpretation errors in a tutorial dialogue system. Allowing students to provide explanations and generate contentful talk can be helpful for learning, but the language that can be understood by a computer system is limited by the current technology. Techniques for dealing with understanding problems have been developed primarily for spoken dialogue syst...

Training task-oriented dialogue systems requires significant amount of manual effort and integration of many independently built components; moreover, the pipeline is prone to errorpropagation. End-to-end training has been proposed to overcome these problems by training the whole system over the utterances of both dialogue parties. In this paper we present an end-to-end spoken dialogue system a...

We present how robustness and adaptivity can be supported by the spoken dialogue system architecture. AthosMail is a multilingual spoken dialogue system for e-mail domain. It is being developed in the EU-funded DUMAS project. It has flexible system architecture supporting multiple components for input interpretation, dialogue management and output generation. In addition to language differences...

In this paper, the influence of intonation to recognize dialogue acts from speech is assessed. Assessment is based on an empirical approach: manually tagged data from a spoken-dialogue and video corpus are used in a CARTstyle machine learning algorithm to produce a predictive model. Our approach involves two general stages: the tagging task, and the development of machine learning experiments. ...

In a human–robot spoken dialogue, the robot may misunderstand an ambiguous command from the user, such as ‘Place the cup down (on the table)’, thus running the risk of an accident. Although asking confirmation questions before the execution of any motion will decrease the risk of such failure, the user will find it more convenient if confirmation questions are not used in trivial situations. Th...