Dialogue Systems Group
Welcome to the Dialogue Systems Group at Cambridge!
The Dialogue Systems group which is part of the CUED Speech Group is led by Prof Steve Young . The work of the group centres on the use of statistical approaches to Spoken Dialogue Systems. The aim is to design systems that can be trained on real dialogue data and which explicitly model the uncertainty present in human-machine interaction. The potential benefits include reduced deployment costs, more robust operation in adverse environments and the ability to adapt on-line. Recent work has primarily focused on the application of Partially-Observable Markov Decision Processes and machine learning techniques to dialogue management and policy design.
Currently the dialogue systems group is working on the EU funded Probabalistic Adaptive Learning And Natural Conversational Engine (PARLANCE) project. The goal of this project is to to design and build mobile applications that approach human performance in con- versational interaction, specifically in terms of the interactional skills needed to do so.
Through the CLASSiC project, we have collected large amount of dialogue data of real user's speech interaction with machines on a tourist information enquiry domain. These data are available for download (free registration) for non-commercial research use.