Landmark-based pronunciation error identification on Chinese learning

Xuesong Yang, Xiang Kong, Mark Allan Hasegawa-Johnson, Yanlu Xie

Research output: Contribution to journalConference article

Abstract

This paper explores a novel approach of identifying pronunciation errors for the second language (L2) learners based on the landmark theory of human speech perception. Earlier works on the selection method of distinctive features and the likelihoodbased “goodness of pronunciation” (GOP) measurement have gained progress in several L2 languages, e.g. Dutch and English. However, the improvement of performance is limited due to error-prone automatic speech recognition (ASR) systems and less distinguishable features. Landmark theory posits the existence of quantal nonlinearities in the articulatory-acoustic relationship, and provides a basis of selecting landmark positions that are suitable for identifying pronunciation errors. By leveraging this English acoustic landmark theory, we propose to select Mandarin Chinese salient phonetic landmarks for the Top-16 frequently mispronounced phonemes by Japanese (L1) learners, and extract features at those landmarks including mel-frequency cepstral coefficients (MFCC) and formants. Both cross validation and evaluation are performed for individual phonemes using support vector machine with linear kernel. Experiments illustrate that our landmark-based approaches achieve higher micro-average f1 score significantly than GOPbased methods.

Original languageEnglish (US)
Pages (from-to)247-251
Number of pages5
JournalProceedings of the International Conference on Speech Prosody
Volume2016-January
StatePublished - Jan 1 2016
Event8th Speech Prosody 2016 - Boston, United States
Duration: May 31 2016Jun 3 2016

Fingerprint

acoustics
learning
language
phonetics
experiment
evaluation
performance
Landmarks
Phoneme
Acoustics

Keywords

  • Acoustic Landmarks
  • Distinctive Features
  • Pronunciation Error Identification
  • Second Language Acquisition

ASJC Scopus subject areas

  • Language and Linguistics
  • Linguistics and Language

Cite this

Landmark-based pronunciation error identification on Chinese learning. / Yang, Xuesong; Kong, Xiang; Hasegawa-Johnson, Mark Allan; Xie, Yanlu.

In: Proceedings of the International Conference on Speech Prosody, Vol. 2016-January, 01.01.2016, p. 247-251.

Research output: Contribution to journalConference article

@article{b75b8da8f16747939d96d86bd8045620,
title = "Landmark-based pronunciation error identification on Chinese learning",
abstract = "This paper explores a novel approach of identifying pronunciation errors for the second language (L2) learners based on the landmark theory of human speech perception. Earlier works on the selection method of distinctive features and the likelihoodbased “goodness of pronunciation” (GOP) measurement have gained progress in several L2 languages, e.g. Dutch and English. However, the improvement of performance is limited due to error-prone automatic speech recognition (ASR) systems and less distinguishable features. Landmark theory posits the existence of quantal nonlinearities in the articulatory-acoustic relationship, and provides a basis of selecting landmark positions that are suitable for identifying pronunciation errors. By leveraging this English acoustic landmark theory, we propose to select Mandarin Chinese salient phonetic landmarks for the Top-16 frequently mispronounced phonemes by Japanese (L1) learners, and extract features at those landmarks including mel-frequency cepstral coefficients (MFCC) and formants. Both cross validation and evaluation are performed for individual phonemes using support vector machine with linear kernel. Experiments illustrate that our landmark-based approaches achieve higher micro-average f1 score significantly than GOPbased methods.",
keywords = "Acoustic Landmarks, Distinctive Features, Pronunciation Error Identification, Second Language Acquisition",
author = "Xuesong Yang and Xiang Kong and Hasegawa-Johnson, {Mark Allan} and Yanlu Xie",
year = "2016",
month = "1",
day = "1",
language = "English (US)",
volume = "2016-January",
pages = "247--251",
journal = "Proceedings of the International Conference on Speech Prosody",
issn = "2333-2042",

}

TY - JOUR

T1 - Landmark-based pronunciation error identification on Chinese learning

AU - Yang, Xuesong

AU - Kong, Xiang

AU - Hasegawa-Johnson, Mark Allan

AU - Xie, Yanlu

PY - 2016/1/1

Y1 - 2016/1/1

N2 - This paper explores a novel approach of identifying pronunciation errors for the second language (L2) learners based on the landmark theory of human speech perception. Earlier works on the selection method of distinctive features and the likelihoodbased “goodness of pronunciation” (GOP) measurement have gained progress in several L2 languages, e.g. Dutch and English. However, the improvement of performance is limited due to error-prone automatic speech recognition (ASR) systems and less distinguishable features. Landmark theory posits the existence of quantal nonlinearities in the articulatory-acoustic relationship, and provides a basis of selecting landmark positions that are suitable for identifying pronunciation errors. By leveraging this English acoustic landmark theory, we propose to select Mandarin Chinese salient phonetic landmarks for the Top-16 frequently mispronounced phonemes by Japanese (L1) learners, and extract features at those landmarks including mel-frequency cepstral coefficients (MFCC) and formants. Both cross validation and evaluation are performed for individual phonemes using support vector machine with linear kernel. Experiments illustrate that our landmark-based approaches achieve higher micro-average f1 score significantly than GOPbased methods.

AB - This paper explores a novel approach of identifying pronunciation errors for the second language (L2) learners based on the landmark theory of human speech perception. Earlier works on the selection method of distinctive features and the likelihoodbased “goodness of pronunciation” (GOP) measurement have gained progress in several L2 languages, e.g. Dutch and English. However, the improvement of performance is limited due to error-prone automatic speech recognition (ASR) systems and less distinguishable features. Landmark theory posits the existence of quantal nonlinearities in the articulatory-acoustic relationship, and provides a basis of selecting landmark positions that are suitable for identifying pronunciation errors. By leveraging this English acoustic landmark theory, we propose to select Mandarin Chinese salient phonetic landmarks for the Top-16 frequently mispronounced phonemes by Japanese (L1) learners, and extract features at those landmarks including mel-frequency cepstral coefficients (MFCC) and formants. Both cross validation and evaluation are performed for individual phonemes using support vector machine with linear kernel. Experiments illustrate that our landmark-based approaches achieve higher micro-average f1 score significantly than GOPbased methods.

KW - Acoustic Landmarks

KW - Distinctive Features

KW - Pronunciation Error Identification

KW - Second Language Acquisition

UR - http://www.scopus.com/inward/record.url?scp=84982965837&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84982965837&partnerID=8YFLogxK

M3 - Conference article

VL - 2016-January

SP - 247

EP - 251

JO - Proceedings of the International Conference on Speech Prosody

JF - Proceedings of the International Conference on Speech Prosody

SN - 2333-2042

ER -