Wednesday, November 2
10:40 AM-11:05 AM
Room 3

Evaluation of Offline Automated Speech Recognition for English as Second Language Learning Application

Full Paper: Research Based ID: 61685
  1. aaa
    Julio Christian Young
    Tokyo Denki University
  2. aaa
    Makoto Shishido
    Tokyo Denki University

Abstract: The second language (L2) acquisition needs are rapidly increasing as the world becomes more globalized. However, learning a new language can be a challenging process and thus requires prolonged commitment and consistent practice. Therefore, several studies in the past have implemented Computer-Assisted Language Learning (CALL) to create a high-impact educational environment that increases students' motivation to learn. Automated Speech Recognition (ASR) is one of the CALL implementations that yields a promising result, especially in learning-by-speaking activities. Despite the great results, most previous ASR studies rely on online ASR technologies, which have several privacy and implementation issues. In this study, we evaluated offline ASR for English as Second Language (ESL) learning by comparing its performance to online ASR and users' perception of its potential use. Based on our early investigation, offline ASR technology could achieve a similar recognition performance to online ASR services by tweaking the recognition process only on specific keywords continuously during the runtime. Moreover, the post-activity questionnaire results showed that the English learning system with offline ASR technology was preferred by the ESL students who participated in this study.

No presider for this session.


Conference attendees are able to comment on papers, view the full text and slides, and attend live presentations. If you are an attendee, please login to get full access.