Dialect classification using vowel acoustic parameters

Citation data:

Speech Communication, ISSN: 0167-6393, Vol: 92, Page: 13-22

Publication Year:
2017
Usage 60
Abstract Views 57
Link-outs 3
Captures 10
Readers 9
Exports-Saves 1
Mentions 1
References 1
Social Media 3
Tweets 3
Citations 1
Citation Indexes 1
DOI:
10.1016/j.specom.2017.05.003
Author(s):
Charalambos Themistocleous
Publisher(s):
Elsevier BV
Tags:
Computer Science; Mathematics; Social Sciences; Arts and Humanities
Most Recent Tweet View All Tweets
article description
This study provides a classification model of two Modern Greek dialects, namely Athenian Greek and Cypriot Greek, using information from formant dynamics of F 1, F 2, F 3, F 4 and vowel duration. To this purpose, a large corpus of vowels from 45 speakers of Athenian Greek and Cypriot Greek was collected. The first four formant frequencies were measured at multiple time points and modelled using second degree polynomials. The measurements were employed in classification experiments, using three classifiers: Linear Discriminant Analysis, Flexible Discriminant Analysis, and C5.0. The latter outperformed the other classification models, resulting in a higher classification accuracy of the dialect. C5.0 classification shows that duration and the zeroth coefficient of F 2, F 3 and F 4 contribute more to the classification of the dialect than the other measurements; it also shows that formant dynamics are important for the classification of dialect.