An open challenge in constructing dialogue systems is developing methods for automatically learning dialogue strategies from large amounts of unlabelled data. Recent work has proposed Next-Utterance-Classification (NUC) as a surrogate task for bui...
A novel online cost-sensitive ensemble learning framework is proposed in this paper, which generalizes a number of batch cost-sensitive ensemble learning algorithms to their online version. Based on this framework, online learning and cost-sensiti...
The authors wish to thank Craig Boutilier, Michael Littman, Andrew Moore and Matthew Mason for many thoughtful comments and discussions regarding this work. We also thank Darius Braziunas, Pascal Poupart, Trey Smith and Nikos Vlassis, for conversa...
Credentials & highlights
- Project Views