Incorporating proximity information for relevance language modeling in speech recognition

Yi Wen Chen, Bo Han Hao, Kuan Yu Chen, Berlin Chen

Research output: Contribution to journalConference article

1 Citation (Scopus)

Abstract

Language modeling (LM), aiming to provide a statistical mechanism to associate quantitative scores to sequences of words, has long been an interesting yet challenging problem in the field of speech and language processing. Although the ngram model remains the predominant one, a number of disparate LM methods have been developed to complement the n-gram model. Among them, relevance modeling (RM) that explores the relevance information inherent between the search history and an upcoming word has shown preliminary promise for dynamic language model adaptation. This paper continues this general line of research in two significant aspects. First, the so-called "bag-of- words" assumption of RM is relaxed by incorporating word proximity evidence into the RM formulation. Second, latent topic information is additionally explored in the hope to further enhance the proximity-based RM framework. A series of experiments conducted on a large vocabulary continuous speech recognition (LVCSR) task seem to demonstrate that the various language models deduced from our framework are very comparable to existing language models.

Original languageEnglish
Pages (from-to)2683-2687
Number of pages5
JournalProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Publication statusPublished - 2013 Jan 1
Event14th Annual Conference of the International Speech Communication Association, INTERSPEECH 2013 - Lyon, France
Duration: 2013 Aug 252013 Aug 29

Fingerprint

Language Modeling
Speech Recognition
Speech recognition
Proximity
Language Model
N-gram
Modeling
Continuous speech recognition
Modeling Method
Dynamic Model
Continue
Complement
Relevance
Series
Formulation
Line
Processing
Model
Demonstrate
Experiment

Keywords

  • Language model
  • Latent topic information
  • Proximity evidence
  • Relevance model
  • Speech recognition

ASJC Scopus subject areas

  • Language and Linguistics
  • Human-Computer Interaction
  • Signal Processing
  • Software
  • Modelling and Simulation

Cite this

Incorporating proximity information for relevance language modeling in speech recognition. / Chen, Yi Wen; Hao, Bo Han; Chen, Kuan Yu; Chen, Berlin.

In: Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH, 01.01.2013, p. 2683-2687.

Research output: Contribution to journalConference article

@article{b7969a8405ea43bebeee3dd28a936b9b,
title = "Incorporating proximity information for relevance language modeling in speech recognition",
abstract = "Language modeling (LM), aiming to provide a statistical mechanism to associate quantitative scores to sequences of words, has long been an interesting yet challenging problem in the field of speech and language processing. Although the ngram model remains the predominant one, a number of disparate LM methods have been developed to complement the n-gram model. Among them, relevance modeling (RM) that explores the relevance information inherent between the search history and an upcoming word has shown preliminary promise for dynamic language model adaptation. This paper continues this general line of research in two significant aspects. First, the so-called {"}bag-of- words{"} assumption of RM is relaxed by incorporating word proximity evidence into the RM formulation. Second, latent topic information is additionally explored in the hope to further enhance the proximity-based RM framework. A series of experiments conducted on a large vocabulary continuous speech recognition (LVCSR) task seem to demonstrate that the various language models deduced from our framework are very comparable to existing language models.",
keywords = "Language model, Latent topic information, Proximity evidence, Relevance model, Speech recognition",
author = "Chen, {Yi Wen} and Hao, {Bo Han} and Chen, {Kuan Yu} and Berlin Chen",
year = "2013",
month = "1",
day = "1",
language = "English",
pages = "2683--2687",
journal = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",
issn = "2308-457X",

}

TY - JOUR

T1 - Incorporating proximity information for relevance language modeling in speech recognition

AU - Chen, Yi Wen

AU - Hao, Bo Han

AU - Chen, Kuan Yu

AU - Chen, Berlin

PY - 2013/1/1

Y1 - 2013/1/1

N2 - Language modeling (LM), aiming to provide a statistical mechanism to associate quantitative scores to sequences of words, has long been an interesting yet challenging problem in the field of speech and language processing. Although the ngram model remains the predominant one, a number of disparate LM methods have been developed to complement the n-gram model. Among them, relevance modeling (RM) that explores the relevance information inherent between the search history and an upcoming word has shown preliminary promise for dynamic language model adaptation. This paper continues this general line of research in two significant aspects. First, the so-called "bag-of- words" assumption of RM is relaxed by incorporating word proximity evidence into the RM formulation. Second, latent topic information is additionally explored in the hope to further enhance the proximity-based RM framework. A series of experiments conducted on a large vocabulary continuous speech recognition (LVCSR) task seem to demonstrate that the various language models deduced from our framework are very comparable to existing language models.

AB - Language modeling (LM), aiming to provide a statistical mechanism to associate quantitative scores to sequences of words, has long been an interesting yet challenging problem in the field of speech and language processing. Although the ngram model remains the predominant one, a number of disparate LM methods have been developed to complement the n-gram model. Among them, relevance modeling (RM) that explores the relevance information inherent between the search history and an upcoming word has shown preliminary promise for dynamic language model adaptation. This paper continues this general line of research in two significant aspects. First, the so-called "bag-of- words" assumption of RM is relaxed by incorporating word proximity evidence into the RM formulation. Second, latent topic information is additionally explored in the hope to further enhance the proximity-based RM framework. A series of experiments conducted on a large vocabulary continuous speech recognition (LVCSR) task seem to demonstrate that the various language models deduced from our framework are very comparable to existing language models.

KW - Language model

KW - Latent topic information

KW - Proximity evidence

KW - Relevance model

KW - Speech recognition

UR - http://www.scopus.com/inward/record.url?scp=84906253884&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84906253884&partnerID=8YFLogxK

M3 - Conference article

AN - SCOPUS:84906253884

SP - 2683

EP - 2687

JO - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

JF - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

SN - 2308-457X

ER -