[논문]자동음성인식 기술을 이용한 모바일 기반 발음 교수법과 영어 학습자의 발음 향상에 관한 연구

박아영

doi:10.9728/dcs.2017.18.6.1101

자동음성인식 기술을 이용한 모바일 기반 발음 교수법과 영어 학습자의 발음 향상에 관한 연구
The Study on Automatic Speech Recognizer Utilizing Mobile Platform on Korean EFL Learners' Pronunciation Development 원문보기

디지털콘텐츠학회 논문지 = Journal of Digital Contents Society, v.18 no.6, 2017년, pp.1101 - 1107

박아영 (창원중앙여자고등학교)

초록
AI-Helper

본 논문은 스마트폰의 플랫폼에 내장되어 있는 자동음성인식 기술을 활용하여 영어 학습자의 발음에 대한 즉각적인 문자 피드백을 제공하는 모바일 기반 발음 교수법이 영어 학습자의 자음 발음 (V-B, R-L, G-Z) 인식과 출력에 미치는 영향에 대해 연구했다. 특히, 자동음성인식 기술을 이용한 모바일 기반 발음 교수법을 사용한 그룹, 전통적인 교사 중심의 발음 교수법 그룹, 그리고 이 둘을 합친 하이브리드 교수법 그룹으로 나누어 영어 학습자의 발음 평가 결과를 (인지, 출력) 비교, 분석했다. ANCOVA를 이용한 분석 결과, 영어 학습자의 발음 출력에 있어 하이브리드 교수법 그룹이 (M=82.71, SD =3.3) 전통적인 교수법 그룹 (M=62.6, SD=4.05) 보다 유의미하게 높은 결과를 나타냈다 (p<.05).

Abstract ▼ AI-Helper

This study explored the effect of ASR-based pronunciation instruction, using a mobile platform, on EFL learners' pronunciation development. Particularly, this quasi-experimental study focused on whether using mobile ASR, which provides voice-to-text feedback, can enhance the perception and production of target English consonants minimal pairs (V-B, R-L, and G-Z) of Korean EFL learners. Three intact classes of 117 Korean university students were assigned to three groups: a) ASR Group: ASR-based pronunciation instruction providing textual feedback by the mobile ASR; b) Conventional Group: conventional face-to-face pronunciation instruction providing individual oral feedback by the instructor; and the c) Hybrid Group: ASR-based pronunciation instruction plus conventional pronunciation instruction. The ANCOVA results showed that the adjusted mean score for pronunciation production post-test on the Hybrid instruction group (M=82.71, SD =3.3) was significantly higher than the Conventional group (M=62.6, SD =4.05) (p<.05).

주제어

AI 본문요약
AI-Helper

* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.

문제 정의

But it is worth noting that there is one difference between [2]’s study and the current studies’ result. That is, the current study included the Hybrid group, which was not usually included in previous ASR studies, and which emphasize the efficiency of the mobile ASR technology.

가설 설정

To make test items suitable to measure target pronunciation, English consonant minimal pairs (V-B, R-L, and G-Z), this study revised and removed items from [1]’s test and included items that aim to test minimal pairs of target pronunciation selected for the current study. The first section of [1]’s test focused on perception of target pronunciation including two questions: 1) Circle the word you hear 2) Circle the word you hear from the minimal pairs. And the second section focused on production of pronunciation including two questions: 1) Say the words 2) Say the minimal pairs.
The first section of [1]’s test focused on perception of target pronunciation including two questions: 1) Circle the word you hear 2) Circle the word you hear from the minimal pairs. And the second section focused on production of pronunciation including two questions: 1) Say the words 2) Say the minimal pairs. The total number of test items for each section was 10.
The results of [13]’s study suggest that ASR was helpful for learners in teaching pronunciation, especially for L2 learners who have limited L2 exposure and strong foreign accents. Therefore, the current study hypothesized that ASR based instruction can be beneficial for L2 pronunciation development of EFL (English as Foreign Language) learners.

제안 방법

Accordingly, this study explored whether ASR using a smartphone platform can be an effective pedagogical tool to enhance the pronunciation teaching and learning of target English pronunciation by comparing it to conventional instruction on EFL learners’ pronunciation development.
As a treatment, the English pronunciation sessions were taken once a week for 20 minutes for one academic semester lasting 15 weeks and were taught by the same instructor. Target pronunciation for this pronunciation sessions were selected before this intervention based on participants’ needs analysis, which identified their learning needs and difficulties.
1, ASR Group received ASR-based pronunciation instruction and were given immediate textual feedback via text message on the smartphone platform. During 20-minute pronunciation sessions, pronunciation training activities consisted of reading aloud the target minimum pairs and phrases in English using the smartphone platform independently. After each reading attempt, students were provided with immediate written text feedback by the smartphone.
Pronunciation instruction consisted of two phases. First, students practiced their target consonants minimal pairs independently using the ASR-based smartphone platform to identify their pronunciation problems. Second, students received face-to-face oral feedback from the instructor focusing on problematic pronunciation identified by the smartphone platform feedback.
Although the Conventional Group did not have access to mobile ASR, they also had to complete the same activities that the ASR Group did. They read aloud the same target English consonants minimal pairs and had weekly 20-minute sessions with an English instructor who provided immediate oral feedback such as recast and repetitions of their pronunciation.
To assess the participants’ perception and production of target pronunciation before and after the intervention, the modified version of [1]’s pronunciation test was adopted.
To maintain reliability of scoring and to minimize rater bias, the participants’ tests were scored by two English instructors who were native speakers of English with considerable experience in grading pronunciation in speaking tests.
To make test items suitable to measure target pronunciation, English consonant minimal pairs (V-B, R-L, and G-Z), this study revised and removed items from [1]’s test and included items that aim to test minimal pairs of target pronunciation selected for the current study.
The total number of test items for each section was 10. To save all the spoken pronunciation performance of participants during the test, mobile ASR on an iPod Touch using a commercial ASR application, Nuance Dragon Dictation, a speaker independent dictation system designed for continuous speech recognition was used. The reason behind is that mobile platform was not an efficient tool to save the large amount of data systematically.

대상 데이터

This experimental study consists of three intact classes of 117 Korean university students which were assigned to three groups: a class of 35 students who received ASR-based instruction, a class of 31 students who received conventional instruction, and a class of 48 students who received Hybrid instruction. All participants were freshmen with an average age of 19 years and 8 months. The three intact classes chosen for the current study were all low-intermediate level classes.
This experimental study consists of three intact classes of 117 Korean university students which were assigned to three groups: a class of 35 students who received ASR-based instruction, a class of 31 students who received conventional instruction, and a class of 48 students who received Hybrid instruction. All participants were freshmen with an average age of 19 years and 8 months.

데이터처리

Secondly, for perception test scores pronunciation, a one-way between-groups analysis of variance (ANOVA) was applied to compare the changes among the three groups between Time 1 and Time 2. However, for productive pronunciation, due to initial difference of among the three groups in the t-test at Time 1, a one-way between-groups analysis of covariance (ANCOVA) was applied to control initial differences between the three groups and for a more statistically accurate comparison among the three groups at Time 2.
The independent variable was the type of treatment (the ASR, Conventional, Hybrid instruction) and the dependent variable was the pronunciation test scores on the pre-test. Secondly, for perception test scores pronunciation, a one-way between-groups analysis of variance (ANOVA) was applied to compare the changes among the three groups between Time 1 and Time 2. However, for productive pronunciation, due to initial difference of among the three groups in the t-test at Time 1, a one-way between-groups analysis of covariance (ANCOVA) was applied to control initial differences between the three groups and for a more statistically accurate comparison among the three groups at Time 2.
To address research question 1 (RQ1), a paired t-test was conducted to examine whether there was a significant difference between pronunciation test scores (perception and production of pronunciation tests respectively) at Time 1 (pre-test) and Time 2 (post-test) for each group (the ASR group, the Conventional group, and the Hybrid group).
To address research question 2 (RQ2), firstly an independent t-test was carried out to examine whether there was an initial difference among the three groups’ pronunciation test scores at Time 1 (pre-test).

성능/효과

One possible explanation for the Hybrid group’s improving the most compared to the other groups was that through Hybrid instruction, ASR-based instruction and Conventional instruction compensate each other and create a synergy effect.
The results of this study should be viewed in light of its limitations, the most obvious of which is the fact that intact classes were used rather than setting up truly experimental condition. From an experimental standpoint, it would be ideal to have a control group in the current study.

후속연구

In addition, because of a short intervention duration and a limited research scope that covered only three selected minimal pairs for the target pronunciation, findings of the current study need to be considered as tentative rather than conclusive. Future research should incorporate a long-term intervention in order to more thoroughly explore the dynamic nature of the impact of the mobile ASR on the wide range of target pronunciation development.

참고문헌 (18)

A. K. Elimat and A. F. AbuSeileek, "Automatic Speech Recognition Technology as an Effective Means for Teaching Pronunciation," JALT CALL Journal, Vol. 10, No. 1, pp. 21-47, 2014.
D. Liakin, W. Cardoso, and N. Liakina, "Learning L2 pronunciation with a mobile speech recognizer: French /y/," CALICO Journal, Vol. 32, No. 1, pp. 12-25, 2015.
J. A. Foote and K. McDonough, "Using shadowing with mobile technology to improve L2 pronunciation," Journal of Second Language Pronunciation, Vol. 3, No. 1, pp. 34-56, 2017.

상세보기
A. Baker, "Exploring teachers' knowledge of second language pronunciation techniques: teacher cognitions, observed classroom practices, and student perceptions," TESOL Quarterly, Vol. 48, No. 1, pp. 136-163, 2014.

상세보기
M. Celce-Murcia, D. Brinton, and J. Goodwin, Teaching pronunciation, 2nd ed. Cambridge: Cambridge University Press, 2010.
S. M. Mccrocklin, "Pronunciation learner autonomy: The potential of Automatic Speech Recognition," System, Vol. 57, pp. 25-42, 2016.

상세보기
M. Li, M. Han, Z. Chen, Y. Mo, X. Chen, and X. Liu, "Improving English Pronunciation Via Automatic Speech Recognition Technology," in Proceeding of Educational Technology (ISET), 2017 International Symposium, pp. 224-228, 2017.
I. S. Lee, "Applications of English education with remote wireless mobile devices," Journal of Digital Contents Society, Vol. 14, No. 2, pp. 255-262, 2013.

원문보기 상세보기
E. J. Song, "A study on the system for on-line education by mobile," Journal of Digital Contents Society, Vol. 6, No. 3, pp. 149-155, 2005.
R. Hincks, "Speech technologies for pronunciation feedback and evaluation," ReCALL, Vol. 15, pp. 3-20, 2003.

상세보기
J. van Doremalen, C. Cucchiarini, and H. Strik, "Automatic pronunciation error detection in non-native speech," Journal of the Acoustical Society of America, Vol. 134, No. 2, pp. 1336-1347, 2013.

상세보기
H. Franco, H. Bratt, R. Rossier, V. Rao Gadde, E. Shriberg, V. Abrash, and K. Precoda, "EduSpeak: A speech recognition and pronunciation scoring toolkit for computer-aided language learning applications," Language Testing, Vol. 27, No. 3, pp. 401-418, 2012.
A. Neri, C. Cucchiarini, H. Strik, and L. Boves, "The pedagogy-technology interface in Computer Assisted Pronunciation Training," Computer Assisted Language Learning, Vol. 15, pp. 441-447, 2002.

상세보기
H. Chen, "Developing and evaluating an oral skills training website supported by automatic speech recognition technology," Recall Journal, Vol. 23, No. 1, pp. 59-78, 2011.

상세보기
C. Gonzalez-Ferreras and V. Cardenoso-Payo, "Improving L2 Production with a Gamified Computer-Assisted Pronunciation Training Tool, TipTopTalk!," in Proceeding of IberSPEECH 2016: IX Jornadas en Tecnologias del Habla and the V Iberian SLTech Workshop, pp. 177-186, 2016.
M. Gutierrez-Colon Plana, P. Gallardo-Torrano, and M. Grova, "SMS as a learning tool: an experimental study," The Eurocall Review, Vol. 20, No. 2, pp. 33-47, 2012.
S. Petersen, R. Sell, and J. Watts, "Let the students lead the way: An exploratory study of mobile language learning in a classroom," in Proceedings 10th World Conference on Mobile and Contextual Learning (mLearn), Beijing, China: Beijing Normal University, pp. 55-61, 2011.
J. M. Howard and A. Scott, "Any time, any place, flexible pace: Technology-enhanced language learning in a teacher education programme," Australian Journal of Teacher Education, Vol. 42, No. 6, pp. 51-68, 2017.

표제어: PCR

동의어: Packet Collision Rate

용어 설명 출처 목록 (6)

용어 설명: PCR은 세균 특이성이 있는 primer를 이용하여 적은 수의 세균이 있을지라도 쉽게 검출할 수 있는 유용한 방법이며, 이를 이용하여 구강 내 치면세균막이나 타액에서 직접 세균을 검출할 수 있게 되었다[8].

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증