Presentation 2009-12-22
Sentence generation from keywords using N-gram for Spoken Dialog System
Yoshitaka YOSHIMI, Yoshihiko NANKAKU, Akinobu LEE, Keiichi TOKUDA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A probabilistic answer selection on spoken dialog system requires a lot of question-and-answer pair as a training data. However, it costs much to collect a sufficient amount of transcription of user's utterance, or to make the question database manually. In this paper, a sentence generation method based on word N-gram to build a question database of a spoken dialog system from given keywords. First, the best word sequences between the keywords are searched based on N-gram. Then, they are concatenated each other and rescored to get the final sentences to be used for the training data. The proposed method was experimented on a public speech-oriented information guidance system. It has been shown that the conventional method with only keywords could get the correct response rate of 59.6%, whereas the proposed method which uses the generated sentences as training data got 65.1%, when using recognized text. Experiments on task-independent N-gram(Web / Newspaper model)also showed that the proposed method can achieve higher accuracy than the conventional method. Comparisons of various parameters were also investigated.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Spoken dialog system / N-gram / Sentence generation / Search
Paper # NLC2009-19,SP2009-83
Date of Issue

Conference Information
Committee NLC
Conference Date 2009/12/14(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Sentence generation from keywords using N-gram for Spoken Dialog System
Sub Title (in English)
Keyword(1) Spoken dialog system
Keyword(2) N-gram
Keyword(3) Sentence generation
Keyword(4) Search
1st Author's Name Yoshitaka YOSHIMI
1st Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology()
2nd Author's Name Yoshihiko NANKAKU
2nd Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology
3rd Author's Name Akinobu LEE
3rd Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology
4th Author's Name Keiichi TOKUDA
4th Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology
Date 2009-12-22
Paper # NLC2009-19,SP2009-83
Volume (vol) vol.109
Number (no) 355
Page pp.pp.-
#Pages 6
Date of Issue