iNETAL

The International English Language Teachers and Lecturers (iNELTAL)

Volume 2022

Research article

The Potential of ASR for Improving English

Pronunciation: A Review
Rahmati Putri Yaniafari*, Viga Olivia, and Suharyadi
Universitas Negeri Malang
ORCID
Rahmati Putri Yaniafari: https://orcid.org/ 0000-0001-8575-6947

Abstract.
To pronounce well is a complex task, requiring students not only to possess knowledge
of the appropriate sounds in a given context, but also to learn to use their vocal
apparatus to make those sounds, equipped with extensive practice and feedback.
Students in these situations require autonomous monitoring experiences to receive
tailored feedback. One of the technological tools learners can use to improve their
pronunciation is Automatic Speech Recognition (ASR). This provides learners with
individual practice and feedback to assist them to accomplish their language goals.
This study examines the database of research on the use of ASR in pronunciation
Corresponding Author: Rahmati instruction and learning available on Google Scholar, Springer Link, Education
Putri Yaniafari; email: Resources Information Center (ERIC), Taylor & Francis Online and Directory of Open
[email protected]
Access Journal (DOAJ). To help the process of identification, some procedures and
Published 28 March 2022 criteria were employed. The results revealed that ten articles met the eligibility criteria.
The procedures of utilizing ASR to improve students’ pronunciation competency are
Publishing services provided by
then discussed in this study.
Knowledge E

Keywords: teaching speaking, teaching pronunciation, Automatic Speech Recognition

Rahmati Putri Yaniafari et
al. This article is distributed
under the terms of the Creative
Commons Attribution License,
which permits unrestricted use
and redistribution provided that 1. Introduction
the original author and source
are credited.
One of the main requirements for students’ language proficiency is to have securing
Selection and Peer-review under pronunciation [1]. This means that students should have an ability to know the sound
the responsibility of the iNETAL
Conference Committee.
of a word which is commonly different from its written. Pronunciation becomes a a
challenging task as it requires students not only to have knowledge of the relevant
sounds in a given context, but also to learn how to pronounce those sounds using
a vocal device, under extensive practice and feedback [2]. Pronunciation is defined
as the act of creating a speech sound that includes intonation articulations, vowels,
accent inflection and formation, in relation to the accuracy or acceptance of utterance
[3]. This complexity is often overlooked as it allows them to deal with phonetic and
phonological competence. Instead, learners emphasize fluency and communication
skills over phonetic accuracy as they have experienced less training on them.

How to cite this article: Rahmati Putri Yaniafari*, Viga Olivia, and Suharyadi, (2022), “The Potential of ASR for Improving English Pronunciation: A
Review” in The International English Language Teachers and Lecturers (iNELTAL), KnE Social Sciences, pages 281–289. DOI 10.18502/kss.v7i7.10670
Page 281
 iNETAL

Under these circumstances, students need independent monitoring experience to

get individualized feedback. This is an alternative way so that they are no longer a reliant
on the teacher who considered as the main facilitator. Language learning appears to
be most efficient when the teacher constantly monitors progress to guide remediation
or advancement [4] . However, in teacher-fronted lessons there is generally not enough
time for sufficient practice and feedback on speaking performance, while traditional
language lab tools usually do not provide the feedback required [5]. Shortly, due to
time constraints in class, providing such training and correcting all student errors is
very difficult to achieve. This encourages the learning process to take advantage of
the technology that has been used this year to assist learning. The rapid development
of technology has opened up an attractive and convenient new arena for language
learning. Particularly, the mobile phone, along with its ubiquitousness and mobility,
offers potential advantages in foreign language learning [6].
With the advancement of information technology, variety of speech tools have
emerged to help learners work on their pronunciation. Automatic Speech Recognition
(ASR) is one of which that help learner reach the language target by offering individual
practice and feedback. ASR is a leading technology that allows humans to interact
with data-processing programs through vocalization [7]. The most advanced systems
incorporating automatic speech recognition (ASR) technology can provide feedback at
the sentence, word, or text level [8]. ASR provides authentic material, such as native
speakers’ pronunciation of the target language, and at the same time allows the
students to listen to and practice their pronunciation in an enjoyable setting; it also
gives each individual learner immediate corrections and feedback, which is difficult to
achieve in class with a large number of students [9]. A stress-free environment promised
by this method encourages students to participate more as autonomous learners. It is
very essential as it could give significant impact on the learners’ pronunciation skill.
Therefore, this study is aimed to review the recent studies of the use of Automatic
Speech Recognition (ASR) in teaching/ learning pronunciation.

2. Recent studies on the Use of ASR to Improve Pronunci-

ation Competence

In the process of identification of relevant studies, a search for articles, conference

papers and thesis was done thoroughly through some research databases, which are
Google Scholar, Springer Link, Education Resources Information Center (ERIC), Taylor
& Francis Online and Directory of Open Access Journal (DOAJ). Several search terms

DOI 10.18502/kss.v7i7.10670 Page 282

 iNETAL

used include “Automatic Speech Recognition”, “Pronunciation”, and “Automatic Speech

Evaluation”, combined with “speaking” and “EFL”. To get the most recent studies, the
articles searched were filtered to those published from 2016 to 2020. The following
criteria of eligibility were also set in view of the research questions.

1. The paper aims at proposing, implementing, or reviewing the use of Automatic

Speech Recognition in teaching/ learning pronunciation

2. The paper is written in English

3. The paper is published between 2016 and 2020

4. The paper is conducted in the context of EFL/ESL

The article identification process (See Figure 1) revealed that there are 10 articles
meeting the criteria of eligibility. The initial search through the aforementioned research
databases shows a great number of papers (n=85). However, after the filtering process,
most of the articles were excluded due to not meeting the requirement of eligibility.

Figure 1: The Process of Articles Identification.

The studies on the use of ASR in EFL setting were conducted on different levels
of education. Out of 11 studies, 2 were investigating the ASR use in EFL classes for a
company employees, 6 were in tertiary level, and 2 others in high school. Table 1 shows
the detail of the summary of the studies.

3. ASR Applications Used

From the 10 studies reviewed, some recommend method in using ASR in teaching
pronunciation. Some of the ten studies that looked into the use of ASR in ESL teaching
and learning employed web-based ASR, while others used software or smartphone

DOI 10.18502/kss.v7i7.10670 Page 283

 iNETAL

Table 1: Summary of the Studies.

No. Author Year Research ASR Used/ The Focus of the

Subject Reviewed Investigation
1. Lara Wallace 2015 University Google Web Proposing an alternative
Student Speech strategy to identify per-
sonal pronunciation prob-
lem using ASR
2 Saandia Ali 2016 Graduate and Various Reviewing existing ASR
Undergraduate Smartphone tools as a preliminary
students Commercial step toward ASR software
Apps and development for CAPT
CAPT
Software
3 Shannon M. 2016 University Windows Investigating the effect of
McCrocklin students joining Speech using ASR on students’
an Advanced Recognition autonomous learning
ESL Listening and voice beliefs and practices
course search on
smart phones
4 Lina Fathi 2017 First year EyeSpeak Investigating the effective-
Sidig Sidgi university Software ness of the ASR soft-
students ware in improving students’
of English pronunciation
Department
5 Xiaobin Liu, 2018 Miidle School Fluent English Trying to see whether the
et al. Students learning app application facilitates the
students’ pronunciation
learning
6 Xiaobin Liu 2019 Freshmen iFlytek Voice Investigating whether
University Input (IVI) the application helps in
Students application improving students’ English
pronunciation
7 Katerina 2020 Employees of “Speechnotes Investigating the impact of
Evers1 and International – Speech to various teaching methods
Sufen Chen Sports Company Text” on the pronunciation per-
joining Business formance of students with
English Course various learning styles are
being investigated.
8 Katerina 2020 Employees of “Speechnotes Investigating whether
Evers & International – Speech to teamwork and feedback
Sufen Chen Sports Company Text” affect learners’ views
joining Business towards ASR
English Course
9 Caiyun Liu 2020 College Not Exploring the phoneme
students mentioned association and phoneme
recognition results in
English
10 Yuting Yuan 2021 3𝑟𝑑 graders “Oral English investigating the effects of
and Xiaobin of Senior Drill & Test” ASR in assisting reading
Liu Highschool App aloud practices on students’
pronunciation

apps, and the remainder did not mention the tool explicitly. The web-based ASR
applications used include Google Web Speech [10] and Speechnotes [11], [12]. When
this article is written, a search on google with “Google Web Speech” key words led to a

DOI 10.18502/kss.v7i7.10670 Page 284

 iNETAL

page called ”Speech to Text by Google,” which offered us a free trial of the service as
well as a price for the premium version. The page also provides a demo that students can
access (See Figure 2). The second web-based ASR application, Speechnotes, also has
both the basic account and the premium offer. When we go to https://speechnotes.co,
however, we can immediately see a clear explanation on how the ASR works, including
how to dictate, punctuate, and proofread (See Figure 3).

Figure 2: The Google Cloud Text-to-Speech Webpage.

Figure 3: The Google Cloud Text-to-Speech Webpage.

The rest of the studies utilized ASR software or smartphone applications, which are
Windows Speech Recognition [13], EyeSpeak Software [9], iFlytek Voice Input [14], Oral
English Drill & Test [15], and Fluent English learning app [16]. Among those software and
applications, only Windows Speech Recognition used by McCrocklin which is accessible
to the author when this paper is written. It can be found in windows computer without
and does not require further installation. The user merely need to verify the settings on
their computer’s ”control panel,” which they may do without having to go online. This

DOI 10.18502/kss.v7i7.10670 Page 285

 iNETAL

feature allows voice commands and text dictation in electronic documents (See Figure
4).

Figure 4: Windows Speech Recognition.

All the ASRs used speech to text procedure, which transcribes any utterances we
make using the connected microphone. EyeSpeak, on the other hand, based on a review
by Tao [17], is designed to facilitate students in learning to produce English phonemes
accurately. It has animated images, sound recordings, graphic articulatory displays, and
record/playback capabilities, according to Tao’s software evaluation (See Figure 5).

Figure 5: EyeSpeak English User Interface [18].

DOI 10.18502/kss.v7i7.10670 Page 286

 iNETAL

4. Proposed Procedure in Using ASR in Teaching Pronun-

ciation

Wallace recommends a method for detecting pronunciation issues using ASR [10]. The
process began with the students recording their speech and ASR transcribing it. The
pupils then rectify the incorrect transcriptions by highlighting them. Next, they re-listen
to their recording and analyze the highlighted words that appear to be mispronounced.
Liu further categorized the pronunciation errors based on the recognition results into
three types: (1) Missing phoneme (phonemes that do not exist in the transcription, but is
silent), (2) Misreading phonemes (phonemes that do not exist in the transcription and not
a missing phoneme), and (3) Added Phonemes (Extra phoneme in the transcription)[2].
However, Evers & Chen’s study of the impact of various teaching methodologies
on the pronunciation performance of students with various learning styles suggests
that ASR feedback has limits in terms of assisting pronunciation improvement [11]. The
findings suggest that combination of ASR feedback and peer-assistance as the most
effective approach compare to the use of ASR alone or teacher’s feedback. Moreover,
the results are applicable for both students with visual and verbal learning style.

5. Conclusion

The analysis of the database shows that the studies on the use of ASR were mostly
conducted in classrooms with adult learners (Out of 11 studies, 2 were investigating the
ASR use in EFL classes for a company employee, 6 were in tertiary level, and 2 others
in high school). The studies revealed that ASR is beneficial in assisting pronunciation
learning. However, although practice using ASR alone is effective [7, 9, 10, 13, 15, 19], a
combination of ASR pronunciation activities with peer instruction may yield even better
results as individual work may frustrate language learners by highlighting the mistakes
in their speech that they cannot fix on their own.

References

[1] Gilakjani AP. A study of factors affecting EFL learners’ English pronunciation learning
and the strategies for instruction. International Journal of Humanities and Social
Science. 2012;2(3):119–128.
[2] McCrocklin SM. Pronunciation learner autonomy: The potential of Automatic Speech
Recognition. System. 2016;57:25–42. https://doi.org/10.1016/j.system.2015.12.013

DOI 10.18502/kss.v7i7.10670 Page 287

 iNETAL

[3] R. M. Dauer, Accurate English: A Complete Course in Pronunciation. Hoboken:

Regents/Prentice Hall, 1993.
[4] Ali S. Towards the development of a comprehensive pedagogical framework for
pronunciation training based on adapted automatic speech recognition systems.
Paper presented at: EURO CALL 2016: Conference on Computer Assisted Language
Learning (CALL); August 24—27, 2016; Lemesos, Cyprus.
[5] van Doremalen J, Boves L, Colpaert J, Cucchiarini C, Strik H. Evaluating
automatic speech recognition-based language learning systems: A
case study. Computer Assisted Language Learning. 2016;29(4):833–851.
https://doi.org/10.1080/09588221.2016.1167090
[6] Youn Ahn T, Lee S-M. User experience of a mobile speaking application with auto-
matic speech recognition for EFL learning: Speaking app with ASR. British Journal
of Educational Technology. 2016;47(4):778–786. https://doi.org/10.1111/bjet.12354
[7] Sidgi LF, Shaari AJ. The usefulness of automatic speech recognition
(ASR) eyespeak software in improving Iraqi EFL students’ pronunciation.
Advances in Language and Literary Studies (ALLS). 2017;8(1):221–226 .
https://doi.org/10.7575/aiac.alls.v.8n.1p.221
[8] Elimat AK, AbuSeileek AF. Automatic speech recognition technology as
an effective means for teaching pronunciation. Japan Association for Lan-
guage Teaching Computer Assisted Language ( JALTCALL). 2014;10(1):21–47.
https://doi.org/10.29140/jaltcall.v10n1.166
[9] Sidgi LFS, Shaari AJ. The effect of automatic speech recognition eye speak software
on Iraqi students’ English pronunciation: A pilot study. Advances in Language and
Literary Studies (ALLS). 2017;8(2):48–54. https://doi.org/10.7575/aiac.alls.v.8n.2p.48
[10] Wallace L. Using Google web speech as a springboard for identifying personal
pronunciation problems. Paper presented at: The 7th Annual Pronunciation in
Second Language Learning and Teaching Conference; October 15—17, 2015; Dallas,
USA.
[11] Evers K, Chen S. Effects of automatic speech recognition software on pronunciation
for adults with different learning styles. Journal of Educational Computing Research.
2020;59(4):669–685. https://doi.org/10.1177/0735633120972011
[12] Evers K, Chen S. Effects of an automatic speech recognition system with peer
feedback on pronunciation instruction for adults. Computer Assisted Language
Learning. 2020;59(4):686–712. https://doi.org/10.1080/09588221.2020.1839504
[13] McCrocklin SM. Pronunciation learner autonomy: The potential of Automatic Speech
Recognition. System. 2016;57:25–42. https://doi.org/10.1016/j.system.2015.12.013

DOI 10.18502/kss.v7i7.10670 Page 288

 iNETAL

[14] Liu X, Xu M, Li M, et al. Improving English pronunciation via automatic speech

recognition technology. International Journal of Innovation and Learning. 2019;25(2):
126—140.
[15] Yuan Y, Liu X. An empirical study of the effect of ASR-supported English reading
aloud practices on pronunciation accuracy. Paper presented at: The 5th International
Conference on Technology in Education (ICTE 2020); August 19–22, 2020; Macau,
China.
[16] Liu X, Zhu C, Jiao J, Xu M. Promoting English pronunciation via mobile devices-
based automatic speech evaluation (ASE) technology. Paper presented at: The 11th
International Conference on Blended Learning (ICBL 2018); 31 July—2 August 2018;
Osaka, Japan.
[17] Tao R. EyeSpeak: Software review. Computer Assisted Language Instruction
Consortium (CALICO) Journal. 2008;25(1):126–136.
[18] Alvaro T. EyeSpeak English software review.Jan. 2009. Accessed on: October 30,
2021. [Online]. Available: https://eyespeak-english.en.uptodown.com/windows.
[19] Liu C. Application of speech recognition technology in pronunciation correction of
college oral English teaching. Paper presented at the 2020 International Conference
on Multi-model Information Analytics (MMIA2020); March 5–6, 2020; Changzhou,
China.

DOI 10.18502/kss.v7i7.10670 Page 289