INTERACTION: Jurnal Pendidikan Bahasa Vol. 10, No.

2: Oktober 2023
P-ISSN: 2406-9558; E-ISSN: 2406-9566

Incorporating AI Tool Along with Traditional method For Speaking

Assessment

Liya Umaroh
[email protected]

Mukaromah
[email protected]

Muhammad Naufal
[email protected]

Ardiawan Bagus Harisa

[email protected]

Universitas Dian Nuswantoro

ABSTRACT
This work focused on incorporating AI tool with traditional method for
speaking assessment. The descriptive qualitative has been implemented to
complete this research. Problem encountered during English pronunciation
was failing to distinguish between short vowel and long vowel. By employing
traditional and AI tool, students got numerous benefits. They may enjoy
learning with the flexibility of time and they also can engage face to face
interaction while using traditional method furthermore, it is essential to
acknowledge AI tool with the limitation and drawbacks. AI tool do not give
the same stage of personal interconnection and on-going response as a human
facilitator.
Key words: AI; Pronunciation; Assessment; Traditional Method

INTRODUCTION

Be in accordance to this education 4.0 English is becoming increasingly important

as a global communication tool, allowing people to work effectively and access
wider resources and information. English is a global lingua franca with 1.45 billion
speakers spread throughout the world. Cultural, social and economic aspects require
English language. It cannot be disavowed that the fundamental role of language is
to convey information from one to another with aim of getting information and
bring the ideas that have been conveyed. Additionally, language is composed of
series of systemized sound, which can be symbols, arbitrary, universal, and
dynamic. (Noermanzah,2022).

Educational technology is a field that focuses on tools and media. Encouraging

communication, knowledge, development and exchange. Currently, 4.0 era

704
demands new capabilities regarding to digital literacy and technological skills, and
enables distance education and online learning through online platforms and
applications. The Edu-tech themselves generate new teaching learning process,
methodology, and varied learning model.

The educational trends have already emerged during 2023 known as artificial
intelligence (AI), online classes, and, VR/AR (Virtual Reality and Augmented
Reality). The development of educational trend will continue to grow at 15.52%
over the next five years in 2027. This is clearly proven by the large number of
studies using AI, Zawacki Richter et al. (2019), Chen et al. (2020a), and Chen et al.
(2022). They focused on AI language education and investigate AI in education in
general (Huang,2023). Another research established by (Chan, 2023), it explored
the AI Ecological Education Policy Framework in higher education with three
different dimensions, they are pedagogical, governance, and operational to
overcome the assorted implication of AI integration universities procedures.
Furthermore, AI is also used to raise students' ability in learning and innovating
language. They can learn through applications, websites, in finding difficult words,
making sentences, improving writing and listening skills (Dwi, 2021). With
numerous AI technologies appearing at the educational level, especially in English
language skills including speaking, listening, writing and grammar, it turns out that
a test is needed to see how optimally AI if it combined with traditional methods.

LITERATURE REVIEW

Previous Related Study

The research conducted by (Noviyanti,2020) focused on the use of artificial

intelligence-based examinations with the initial aim of seeing how considerable the
impact of mastering pronunciation skills seen from the perspective of students
majoring in IAIN Jember, the research had implemented mixed qualitative and
quantitative research methods. Data collection begins by providing pretest activities
and ends with a post test. The learning results show that the score increased by 30
points from the post test score of 56 to 90 when the post test was carried out. The
results of the questionnaire showed how to learn faster and help learn pronunciation
independently.

The second research was conducted by (Chen, 2022). It described This research has
associated both teachers and students with the main purpose of seeing the impact
of AI technology in the English learning process. The stages carried out with giving
an oral presentation assignment in English language, then the teacher focuses
attention during the presentation process and last, it uses AI to provide responses
when students make a presentation. The results conveyed that student felt happy
when assists AI as a learning medium. However, there is another aspect that is
lacking when using AI during presentations, pronunciation, and acuraione which is
different from traditional methods.

705
Artificial Intelligence in Current Education

AI is an abbreviation of "Artificial Intelligence" which is a field in computer science

that focuses on developing computer systems that can perform tasks that normally
require human intelligence. The main goal of AI is to develop computers or
machines that can learn, think and behave like humans. In general, artificial
intelligence was created to help humans solve problems. Various applications have
been designed to optimize workload and ultimately reduce the role of a complete
human being (Huang, 2023)

Speech Recognition Through Artificial Intelligence and Pronunciation

Speech recognition is one of the emerging areas of machine intelligence growing

rapidly, this is indicated by almost all technological devices being equipped by
voice command. This has been interesting for researchers to make speech
recognition as an important discipline to create impact technology on society and is
expected to develop further in the field machine interaction with humans. This
technology allows a device to recognize and understand spoken words by digitizing
the words and matching the digital signal with a certain pattern stored in a device.
Spoken words are transformed into digital signals by converting sound waves into
a set of numbers which are then adjusted to certain codes to identify the words. The
results of identifying spoken words can be displayed in written form or can be read
by a technological device as a command to carry out a job, for example pressing a
button on a cell phone which is done automatically with a voice command.
(Kazu,2023)

Pronunciation

Pronunciation is a set of habits of producing sounds that is acquired by repeating

over and over again by being corrected when it is pronounced wrongly. It includes
attention to the particular sounds of language, aspect of speech such as intonation,
phrasing, stress, timing, rhythm that are closely to the way we speak language.
(Fraser, 2001:16)

METHOD

This research implements descriptive qualitative. Combining With AI speech

recognition using web devices and traditional method as the instrument of this
research. Here is the research design:

1. Data Collection

2. Data Analysis

706
 3. Preparation report

4. conclusion/objective
result

Figure. 1 Research design

The data analysis is going to describe with explanation together with chart. The
sample of this research were student from higher education who joined in roller
community around 25 students from different university. Pronunciation assessment
will be given both using speech recognition with AI tool and also traditional
method.

RESULT AND DISUSSION

Common Problem Encountered in English Pronunciation

From depth interview with 25 students during four months. It found some common
English pronunciation problem as follows:

Common Pronunciation Problem faced by learner

1) The different between
sound
2) The different between
vowel sound
3) The Schwa and it’s problem
16% 13%
4) Difficult sound for some

12% 19% 5) Another language another

sound
6) Sentence stress
12% 9%
9% 10% 7) Intonation

8) Connected speech

Figure 2 Common Pronunciation Problem

707
The interview delivered with open ended questions and the result describe on the
pie chart above. The common problem that often face by English learner are: 1) the
different between sound 2) the different between vowel sound 3) the schwa and its
problem 4) difficult sound for some 5) another language another sound 6) sentence
stress 7) intonation and 8) connected speech. From the eight problem the different
between vowel sound reached the highest 19% as the big problem in pronunciation.
By this result the next step for this data analysis will be focused on the different
between vowel sound.

English Vowel Sound (short vs. long vowel)

In English language, five letters among twenty-six letters of English alphabet are
called vowels. They are (a, i, u, e, o, and u) and sometimes ‘y’ is also called vowel.
These are representing vowel sound

English short vowel : /ɪ/, /e/, /æ/, /ʌ/, /ʊ/ , /ɒ/, /ə/
/ɪ/ fit /fiːt/
/e/ pet /pet/
æ/ pat /pæt/
/ʌ/ cut /kʌt/
/ʊ/ put /pʊt/,
/ɒ/ pot /pɒt/
/ə/ about /əˈbaʊt/

English long vowel :/i:/, /ɑ:/, /ɔ:/, /ɜ:/, /u:/.

/i:/ week /wi:k/
/ɑ:/ hard /ha:/
/ɔ:/ fork /fɔ:k/
/ɜ:/ heard /hɜ:d/
/u:/ boot /bu:t/

Pronunciation Assessment using traditional rubric/traditional method

This assessment implements with four different vowel word, they are:

Table.1 four words long and short vowel for assessment

blend blɛnd/
trap /Træp/
hard /ha:d/
mud / mʌd/
Table. 2 rubrics

Vowel sound Poor Fair Excellent

708
 Recognizing long the student was The student The student
vowel sound not able to recognizes 1 long recognizes 2 long
recognize long vowel sound vowel sound
vowel sound
Recognizing short the student was The student The student
vowel sound not able to recognizes 1 short recognizes 2 short
recognize short vowel sound vowel sound
vowel sound

Here is the result of assessing pronunciation with rubric/traditional method

Result of pronunciation with traditional method

20

15

10

5 excellent

0 fair
Blend
Trap poor
Hard
Mud

poor fair excellent

Reading the result from the chart above most of students have middle fluency in
pronouncing long and short vowel.

Pronunciation Assessment using AI tool

After all the students practicing the four words (hard, mud, blend, trap) and have
already done the assessment with traditional method. They continue to use AI tool,
where it was developed by co-researcher from informatic science. Here is the
performance of pronunciation AI tool website, link:
https://violance.risetkami.com/pronunciation

709
 Figure 4. AI tool for assessing pronunciation

Here are the steps for using this AI tool:

Step 1 - Choose One Words
Choose one of the words from (BLEND, TRAP, HARD, OR MUD)
Step 2 - Press the Record Button
Click on the "Start Recording" button.
Step 3 - Allow Microphone Access & Start Talking
A popup will appear, asking you permission for our website to access your
microphone. Click on "Allow" to start recording. Speak clearly into your
microphone for up to 5 minutes.
Step 3 - Stop Recording
Recording will automatically stop when you reach the duration limit, or you can
stop it anytime by clicking on the "Stop" button.
Step 4 - Preview your Recording or Try Again
Once you finish recording, an audio player will appear. Press the play button to
preview your recording, or click "Redo Recording" to try again.
Step 5 - Download Your Audio Recording
If you are happy with your final recording, click the "Download Audio" button.
Step 6 - Press Predict Button
To predict your Pronunciation score, click the "Predict" button.
Step 7 - The Scores
Your Pronunciation score will be shown in bottom area
Before trying to click the word, it provides explanation about short and long vowel
along with the examples. Here is the result from 25 students attempting four words
(blend, trap, hard, mud)

710
 Correct and incorrect pronunciation with AI

80

60

40

20

0 incorrect
hard
mud correct
trap
blend

correct incorrect

Compare to the traditional assessment, the AI tool showed different result. It is

because A lot of advantages in applying AI for language learning but it is essential
to acknowledge the limitation and drawbacks. AI tool do not give the same stage of
personal interconnection and on-going response as a human facilitator.

CONCLUSION

By incorporating AI tool and traditional method for assessing pronunciation,

students got numerous benefits. They may enjoy learning with the flexibility of time
and they also can engage face to face interaction while using traditional method.

REFERENCES

Abdulmunem, R. A. (2023). Artificial intelligence in education. In Comparative

Research on Diversity in Virtual Learning: Eastern vs. Western Perspectives.
https://doi.org/10.4018/978-1-6684-3595-3.ch012
Chan, C. K. Y. (2023). A comprehensive AI policy education framework for
university teaching and learning. International Journal of Educational
Technology in Higher Education, 20(1). https://doi.org/10.1186/s41239-023-
00408-3
Chen, L., Chen, P., & Lin, Z. (2020). Artificial Intelligence in Education: A Review.
IEEE Access, 8, 75264–75278.
https://doi.org/10.1109/ACCESS.2020.2988510
Dewi, H. K., Rahim, N. A., Putri, R. E., Wardani, T. I., & Pandin, M. G. R. (2021).
The use of AI (Artificial Intelligent) in English learning among university
student: A case study in English department, Universitas Airlangga. Artikel
Ilmiah Filsafat, Ml, 11–10.

711
Dewi, H. K., Rahim, N. A., Putri, R. E., Wardani, T. I., Rumambo, M. G.,
Airlangga, U., Glorino, M., Pandin, R., & Airlangga, U. (n.d.). Artikel ilmiah
filsafat final. Ml.
Ginther, A. (2012). O ller, J ohn W ., J r. . The Encyclopedia of Applied Linguistics,
November. https://doi.org/10.1002/9781405198431.wbeal0882
Harahap, N., & Jupriaman. (2019). Upaya Peningkatan Pronunciation melalui
Ragam Metode Student Centered Learning (SCL) dan Karakter di Prodi
Pendidikan Bahasa Inggris. RED Revolution of English Department Journal,
3(2), 1–7.
Kazu, İ. Y., & Kuvvetli, M. (2023). The Influence of Pronunciation Education via
Artificial Intelligence Technology on Vocabulary Acquisition in Learning
English. International Journal of Psychology and Educational Studies, 10(2),
480–493. https://doi.org/10.52380/ijpes.2023.10.2.1044
Liang, Y., Song, K., Mao, S., Jiang, H., Qiu, L., Yang, Y., Li, D., Xu, L., & Qiu, L.
(2023). End-to-End Word-Level Pronunciation Assessment with MASK Pre-
training. 969–973. https://doi.org/10.21437/interspeech.2023-585
Nguyen, H. D. T. (2018). Finding proper pronunciation assessment. Ho Chi Minh
City Open University Journal of Science - Social Sciences, 8(1), 40–47.
https://doi.org/10.46223/hcmcoujs.soci.en.8.2.286.2018
Noviyanti, S. D. (2020). Artificial Intelligence (AI)-Based Pronunciation Checker:
An Alternative for Independent Learning in Pandemic Situation. ELT Echo :
The Journal of English Language Teaching in Foreign Language Context,
5(2), 162. https://doi.org/10.24235/eltecho.v5i2.7246
Purnawarman, P., & Darajati, U. (2020). Authentic Speaking Assessment Applied
by English Teachers During Online Learning. 509(Icollite), 560–566.
https://doi.org/10.2991/assehr.k.201215.087
Sabata, Y. N., & Liise, D. (n.d.). The Effect of Mother Tongue toward English
Pronunciation. 1(2), 73–79.
Suciati, S., Faridi, A., Mujiyanto, J., & Arifani, Y. (2023). Artificial Intelligence
Application dalam Pembelajaran Speaking: Persepsi dan Solusi. Prosiding
Seminar Nasional Pascasarjana, 1111–1115.
http://pps.unnes.ac.id/pps2/prodi/prosiding-pascasarjana-unnes
Wen, D., & Lin, F. (2008). Ways and means of employing AI technology in e-
learning systems. Proceedings - The 8th IEEE International Conference on
Advanced Learning Technologies, ICALT 2008, January, 1005–1006.
https://doi.org/10.1109/ICALT.2008.304
Yuan, C. (2023). On the Transfer Effect of Sichuan Dialect to the Learning of
English Pronunciation. The Educational Review, USA, 7(3), 329–332.
https://doi.org/10.26855/er.2023.03.008
Zakiyyah, F., Setyaji, A., & Ardini, S. N. (2022). The Analysis of Pronunciation
Application Based on the Concept of Artificial Intelligence. 2(1), 559–569.

712