Computational prosodic features analysis of children's Filipino speech for automated oral reading fluency assessment

College

College of Computer Studies

Document Type

Conference Proceeding

Source Title

Philippine Computing Science Congress 2020

Publication Date

3-2020

Abstract

This paper explores the computational extraction and analysis of prosodic features in children’s Filipino speech for application in automated oral reading fluency assessment. Automatic syllabication was optimized on children’s Filipino speech using the Children Filipino Speech Corpus. Prosodic features such as speech rate, articulation rate, pauses, syllable duration, and pitch were extracted and grouped according to human rater assessment of fluency. ANOVA results showed that speech rate, articulation rate, pauses, syllable duration, and to some extent, pitch, as measures of prosody, can be used to predict children’s oral reading fluency in Filipino.

html

Disciplines

Language and Literacy Education

Keywords

Oral reading; Fluency (Language learning); Filipino language—Versification; Audio data mining

Upload File

wf_no

This document is currently not available here.

Share

COinS