Study on Ambiguous Understanding Representation and Response Generation for Spoken Dialog Systems
Project/Area Number |
19700164
|
Research Category |
Grant-in-Aid for Young Scientists (B)
|
Allocation Type | Single-year Grants |
Research Field |
Perception information processing/Intelligent robotics
|
Research Institution | Nagoya University |
Principal Investigator |
KITAOKA Norihide Nagoya University, 大学院・情報科学研究科, 准教授 (10333501)
|
Project Period (FY) |
2007 – 2009
|
Project Status |
Completed (Fiscal Year 2009)
|
Budget Amount *help |
¥3,570,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥570,000)
Fiscal Year 2009: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2008: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2007: ¥1,100,000 (Direct Cost: ¥1,100,000)
|
Keywords | 音声対話 / 誤認識 / 複数理解状態 / 効率性 / 無矛盾性 / グラフ探索 / ヒューリスティック探索 / ユーザ満足度 / 音声対話システム / 応答戦略 / 自然性 |
Research Abstract |
We proposed a new dialog strategy which can recover from system's misunderstandings without confirmation system utterances. In this strategy, system uses a graph search algorithm, in which each node represent an understanding state. To make a response to the user, system adopts 'efficiency measure' which enables the system efficient search and 'consistency measure' which prevents unnatural response inconsistent with the dialog history. We developed a system with these methods and evaluated it. It is proved that the system was more efficient and obtained more user satisfaction than conventional ones.
|
Report
(4 results)
Research Products
(10 results)