Computer-assisted pronunciation training in Icelandic (CAPTinI): developing a method for quantifying mispronunciation in L2 speech-Reference-Cited by-同舟云学术

Computer-assisted pronunciation training in Icelandic (CAPTinI): developing a method for quantifying mispronunciation in L2 speech

Published:2022-12-12 Issue: Volume: Page:334-339
ISSN:
Container-title:Intelligent CALL, granular systems and learner data: short papers from EUROCALL 2022
language:
Short-container-title:

Author:

Richter Catlin¹^ORCID,Bédi Branislav²^ORCID,Pálsson Ragnar¹,Guðnason Jón¹^ORCID

Affiliation:

1. Reykjavík University, Reykjavík, Iceland

2. The Árni Magnússon Institute for Icelandic Studies, Reykjavík, Iceland

Abstract

We are developing a Computer-Assisted Pronunciation Training (CAPT) system for learners of Icelandic as a second and foreign language (L2). Based on pre-designed tasks in pronunciation exercises, this system will provide corrective feedback on learners’ speech. One of the main features we are implementing is a new method for automatic pronunciation scoring to provide immediate feedback on learners’ errors. We report promising results for a pilot study of this method in Norwegian, where the pronunciation score successfully distinguishes between native speakers and adult learners, and we discuss how this method informs our continued development of Icelandic CAPT.

Publisher

Research-publishing.net

Link

https://research-publishing.net/publication/chapters/978-2-38372-015-7/1480.pdf

Reference8 articles.

1. Icelandic Online: twenty years of development, evaluation, and expansion of an LMOOC

2. A New Acoustic-Based Pronunciation Distance Measure

3. Bédi, B. (2022). Development of online tools supporting the learning of Icelandic as a foreign and second language. In Tungumál í víðu samhengi: Afmælisrit til heiðurs Birnu Arnbjörnsdóttur, (pp. 47-56). Reykjavík.

4. The Quality of Language Learning Opportunities

5. Automatic assessment of English proficiency for Japanese learners without reference sentences based on deep neural network acoustic models