Abstract | ||
---|---|---|
In real-world applications, modelling dialogue as a POMDP requires the use of a summary space for the dialogue state representation to ensure tractability. Sub-optimal estimation of the value function governing the selection of system responses can then be obtained using a grid-based approach on the belief space. In this work, the Monte-Carlo control technique is extended so as to reduce training over-fitting and to improve robustness to semantic noise in the user input. This technique uses a database of belief vector prototypes to choose the optimal system action. A locally weighted k-nearest neighbor scheme is introduced to smooth the decision process by interpolating the value function, resulting in higher user simulation performance. |
Year | Venue | Keywords |
---|---|---|
2009 | SIGDIAL Conference | dialogue state representation,system response,belief space,monte-carlo control technique,summary space,pomdp-based dialogue system,optimal system action,belief vector prototype,value function,k-nearest neighbor monte-carlo control,modelling dialogue,higher user simulation performance,monte carlo,k nearest neighbor,optimal estimation |
Field | DocType | Citations |
k-nearest neighbors algorithm,Monte Carlo method,Computer science,Partially observable Markov decision process,Interpolation,Bellman equation,Robustness (computer science),Artificial intelligence,Communication noise,Machine learning,Grid | Conference | 3 |
PageRank | References | Authors |
0.44 | 7 | 8 |
Name | Order | Citations | PageRank |
---|---|---|---|
Fabrice Lefèvre | 1 | 185 | 26.62 |
M. Gašić | 2 | 38 | 2.40 |
F. Jurčíček | 3 | 35 | 1.90 |
S. Keizer | 4 | 35 | 1.90 |
F. Mairesse | 5 | 84 | 5.44 |
B. Thomson | 6 | 55 | 3.63 |
Kai Yu | 7 | 1082 | 90.58 |
S. Young | 8 | 134 | 7.84 |