Policy optimisation of POMDP-based dialogue systems without state space compression

M. Gasic, M. Henderson, B. Thomson, P. Tsiakoulis, S. Young
2012 2012 IEEE Spoken Language Technology Workshop (SLT)  
The partially observable Markov decision process (POMDP) has been proposed as a dialogue model that enables automatic improvement of the dialogue policy and robustness to speech understanding errors. It requires, however, a large number of dialogues to train the dialogue policy. Gaussian processes (GP) have recently been applied to POMDP dialogue management optimisation showing an ability to substantially increase the speed of learning. Here, we investigate this further using the Bayesian
more » ... of Dialogue State dialogue manager. We show that it is possible to apply Gaussian processes directly to the belief state, removing the need for a parametric policy representation. In addition, the resulting policy learns significantly faster while maintaining operational performance.
doi:10.1109/slt.2012.6424165 dblp:conf/slt/GasicHTTY12 fatcat:y5nryd22kjfv5ih24ullxtv4hy