Automatic scoring of pronunciation quality with hybrid measure

Bin Dong; Fengpei Ge; Fuping Pan; Shui Duen Chan

doi:10.1109/AICI.2009.404

Automatic scoring of pronunciation quality with hybrid measure

Bin Dong, Fengpei Ge, Fuping Pan, Shui Duen Chan

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

1 Citation (Scopus)

Abstract

In this paper, a hybrid measure for automatic scoring of Mandarin pronunciation quality is presented. Different to prevalent algorithms, mono-phone-based and tri-phone-based acoustic models are applied and two types of features are combined to get the score of "Goodness of Pronunciations" with Support Vector Machine algorithm, which are the average logarithm of the frame-based posterior probability and the normalized logarithm of the phoneme-based posterior probability. With the hybrid measure, the average correlation coefficient between machine scores from automatic system and the rater scores is improved from 0.8379 to 0.8549 which almost reach the coefficient 0.8564 between different raters.

Original language	English
Title of host publication	2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009
Pages	381-384
Number of pages	4
Volume	3
DOIs	https://doi.org/10.1109/AICI.2009.404
Publication status	Published - 1 Dec 2009
Event	2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009 - Shanghai, China Duration: 7 Nov 2009 → 8 Nov 2009

Conference

Conference	2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009
Country/Territory	China
City	Shanghai
Period	7/11/09 → 8/11/09

Keywords

Automatic scoring
Hybrid meansure
Speech recognition

ASJC Scopus subject areas

Artificial Intelligence
Computational Theory and Mathematics
Computer Graphics and Computer-Aided Design
Software

More information

10.1109/AICI.2009.404

Cite this

@inproceedings{d35a1d1173514c13b473722ceb8f4f65,

title = "Automatic scoring of pronunciation quality with hybrid measure",

abstract = "In this paper, a hybrid measure for automatic scoring of Mandarin pronunciation quality is presented. Different to prevalent algorithms, mono-phone-based and tri-phone-based acoustic models are applied and two types of features are combined to get the score of {"}Goodness of Pronunciations{"} with Support Vector Machine algorithm, which are the average logarithm of the frame-based posterior probability and the normalized logarithm of the phoneme-based posterior probability. With the hybrid measure, the average correlation coefficient between machine scores from automatic system and the rater scores is improved from 0.8379 to 0.8549 which almost reach the coefficient 0.8564 between different raters.",

keywords = "Automatic scoring, Hybrid meansure, Speech recognition",

author = "Bin Dong and Fengpei Ge and Fuping Pan and Chan, {Shui Duen}",

year = "2009",

month = dec,

day = "1",

doi = "10.1109/AICI.2009.404",

language = "English",

isbn = "9780769538167",

volume = "3",

pages = "381--384",

booktitle = "2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009",

note = "2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009 ; Conference date: 07-11-2009 Through 08-11-2009",

}

Dong, B, Ge, F, Pan, F & Chan, SD 2009, Automatic scoring of pronunciation quality with hybrid measure. in 2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009. vol. 3, 5376710, pp. 381-384, 2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009, Shanghai, China, 7/11/09. https://doi.org/10.1109/AICI.2009.404

Automatic scoring of pronunciation quality with hybrid measure. / Dong, Bin; Ge, Fengpei; Pan, Fuping et al.
2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009. Vol. 3 2009. p. 381-384 5376710.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - Automatic scoring of pronunciation quality with hybrid measure

AU - Dong, Bin

AU - Ge, Fengpei

AU - Pan, Fuping

AU - Chan, Shui Duen

PY - 2009/12/1

Y1 - 2009/12/1

N2 - In this paper, a hybrid measure for automatic scoring of Mandarin pronunciation quality is presented. Different to prevalent algorithms, mono-phone-based and tri-phone-based acoustic models are applied and two types of features are combined to get the score of "Goodness of Pronunciations" with Support Vector Machine algorithm, which are the average logarithm of the frame-based posterior probability and the normalized logarithm of the phoneme-based posterior probability. With the hybrid measure, the average correlation coefficient between machine scores from automatic system and the rater scores is improved from 0.8379 to 0.8549 which almost reach the coefficient 0.8564 between different raters.

AB - In this paper, a hybrid measure for automatic scoring of Mandarin pronunciation quality is presented. Different to prevalent algorithms, mono-phone-based and tri-phone-based acoustic models are applied and two types of features are combined to get the score of "Goodness of Pronunciations" with Support Vector Machine algorithm, which are the average logarithm of the frame-based posterior probability and the normalized logarithm of the phoneme-based posterior probability. With the hybrid measure, the average correlation coefficient between machine scores from automatic system and the rater scores is improved from 0.8379 to 0.8549 which almost reach the coefficient 0.8564 between different raters.

KW - Automatic scoring

KW - Hybrid meansure

KW - Speech recognition

UR - http://www.scopus.com/inward/record.url?scp=77949293872&partnerID=8YFLogxK

U2 - 10.1109/AICI.2009.404

DO - 10.1109/AICI.2009.404

M3 - Conference article published in proceeding or book

SN - 9780769538167

VL - 3

SP - 381

EP - 384

BT - 2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009

T2 - 2009 International Conference on Artificial Intelligence and Computational Intelligence, AICI 2009

Y2 - 7 November 2009 through 8 November 2009

ER -

Automatic scoring of pronunciation quality with hybrid measure

Abstract

Conference

Keywords

ASJC Scopus subject areas

More information

Other files and links

Fingerprint

Cite this