Presentation | 2015-10-16 Evaluation of Parallel POMDP based on State Space Partition and State Transition Learning Yoji Yamada, Tetsuya Takiguchi, Yasuo Ariki, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Significant problems with Partially Observable Markov Decision Process (POMDP) lie in a large number of states, causing a long computation time and difficulty of system development. In order to solve these problems, we have applied Hierarchical POMDP (H-POMDP) to the spoken dialogue system (SDS) so far. However, in H-POMDP, the transitions between the states are restricted because of the partition of state space and the hierarchical structure. In this paper, we propose Parallel POMDP (P-POMDP) in which small POMDPs are scattered and they work together in parallel to relax this restriction. In P-POMDP, the reward function is newly formalized to make the system reply optimally. In the experiment, we compared the proposed P-POMDP with H-POMDP and the conventional POMDP in terms of task achievement rates and computation times. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | POMDP / Spoken Dialogue System / Reward Function / State Space |
Paper # | SP2015-66 |
Date of Issue | 2015-10-08 (SP) |
Conference Information | |
Committee | SP |
---|---|
Conference Date | 2015/10/15(2days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | Kobe Univ. |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | Speech interface, Synthesis, Dialogue, Application system, etc. |
Chair | Kazunori Mano(Shibaura Inst. of Tech.) |
Vice Chair | Norihide Kitaoka(Tokushima Univ.) |
Secretary | Norihide Kitaoka(Tokyo City Univ.) |
Assistant | Takashi Nose(Tohoku Univ.) / Taichi Asami(NTT) |
Paper Information | |
Registration To | Technical Committee on Speech |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Evaluation of Parallel POMDP based on State Space Partition and State Transition Learning |
Sub Title (in English) | |
Keyword(1) | POMDP |
Keyword(2) | Spoken Dialogue System |
Keyword(3) | Reward Function |
Keyword(4) | State Space |
1st Author's Name | Yoji Yamada |
1st Author's Affiliation | Kobe University(Kobe Univ.) |
2nd Author's Name | Tetsuya Takiguchi |
2nd Author's Affiliation | Kobe University(Kobe Univ.) |
3rd Author's Name | Yasuo Ariki |
3rd Author's Affiliation | Kobe University(Kobe Univ.) |
Date | 2015-10-16 |
Paper # | SP2015-66 |
Volume (vol) | vol.115 |
Number (no) | SP-253 |
Page | pp.pp.39-43(SP), |
#Pages | 5 |
Date of Issue | 2015-10-08 (SP) |