Presentation | 1993/12/10 Topic Focusing Mechanism for Speech Recognition based on Probabilistic Grammar and Topic Markov Model Takeshi Kawabata, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Speech conversation is the most comfortable interface for human- machine communication.For this purpose,effective natural language processing technologies are necessary.This paper describes a new stochastic topic focusing mechanism for reducing the perplexity of natural spoken languages.The predictive CFG parser analyzes input speech and generates grammar-rule sequences.These rule sequences drive the topic hidden Markov model and the current topic is estimated as HMM state distribution.The CFG rule probabilities are dynamically changed according to this topic state distribution.The evaluation experiments using a large dialog text database shows that the proposed topic focusing mechanism effectively reduce the task perplexity. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Natural language / Topic focusing / Speech understanding / JUNO |
Paper # | NLC93-55 |
Date of Issue |
Conference Information | |
Committee | NLC |
---|---|
Conference Date | 1993/12/10(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) | Topic Focusing Mechanism for Speech Recognition based on Probabilistic Grammar and Topic Markov Model |
Sub Title (in English) | |
Keyword(1) | Natural language |
Keyword(2) | Topic focusing |
Keyword(3) | Speech understanding |
Keyword(4) | JUNO |
1st Author's Name | Takeshi Kawabata |
1st Author's Affiliation | NTT Basic Research Laboratories() |
Date | 1993/12/10 |
Paper # | NLC93-55 |
Volume (vol) | vol.93 |
Number (no) | 367 |
Page | pp.pp.- |
#Pages | 7 |
Date of Issue |