ISSN: 3030-3931, Impact factor: 7,241
Volume 8, issue2, Iyul 2025
https://worldlyjournals.com/index.php/Yangiizlanuvchi
worldly knowledge
OAK Index bazalari :
research gate, research bib.
Qo’shimcha index bazalari:
zenodo, open aire. google scholar.
Original article
197
THE ROLE OF SPEECH RECOGNITION AI IN ENHANCING LISTENING AND
PRONUNCIATION FOR UZBEK LEARNERS
Baxramova Malika Muzaffarovna
Urgench State Pedagogical Institute
Abstract:
The integration of speech recognition technology into English as a Foreign Language
(EFL) instruction has opened new possibilities for improving learners’ listening comprehension
and pronunciation. This study investigates the potential of Artificial Intelligence (AI)-driven
speech recognition tools to support Uzbek learners in developing their auditory and oral
language skills. Drawing on current practices, learner needs, and the pedagogical limitations of
traditional classrooms, the article explores how AI applications such as Duolingo, Elsa Speak,
and Google’s Speech-to-Text contribute to more autonomous, personalized, and effective
language acquisition. These tools provide real-time corrective feedback, interactive listening
input, and pronunciation assessment tailored to individual learners. Uzbek students, often
challenged by the phonological differences between English and Uzbek, benefit from repeated
exposure, accent variation, and targeted error correction. Despite significant advantages, the
implementation of such tools in Uzbekistan faces challenges, including unequal access to
technology, limited teacher training, and language interface barriers. The study concludes that,
with proper infrastructural and methodological support, speech recognition AI can serve as a
valuable aid in enhancing both listening and speaking outcomes in the Uzbek EFL context.
Keywords :
Speech recognition, Artificial Intelligence, listening comprehension, pronunciation,
Uzbek EFL learners, language technology, educational apps, autonomous learning, phonological
awareness, digital pedagogy.
The advancement of Artificial Intelligence (AI) technologies has brought significant innovation
to the field of language education, particularly through the development of speech recognition
systems. These tools, which analyze and respond to spoken language input in real time, offer
valuable support for enhancing both listening comprehension and pronunciation skills. For
Uzbek learners of English as a Foreign Language (EFL), who often struggle with limited
exposure to authentic spoken English and lack sufficient speaking practice opportunities, AI-
powered speech recognition presents a practical and scalable solution. This article explores the
role of speech recognition AI in improving listening and pronunciation among Uzbek learners,
emphasizing its pedagogical potential, learner engagement, and challenges in implementation.
Speech recognition AI operates by converting a learner’s spoken input into text and comparing it
with a model of native speaker output. The system then provides feedback on various dimensions,
such as pronunciation accuracy, fluency, intonation, and word stress. In listening activities, these
tools assist learners by providing synchronized subtitles, playback control, and comprehension
questions based on the spoken input. When paired together, these listening and speaking
components form a loop of interactive learning that allows students to hear accurate language
input and immediately practice reproducing it.
For many Uzbek EFL learners, especially those in public schools or rural areas, traditional
English instruction may rely heavily on grammar and translation, with limited opportunities for
oral practice. In such environments, speech recognition tools like Google's Speech-to-Text,
Microsoft's Azure, or app-based platforms such as Elsa Speak, Duolingo, or Cake can provide a
ISSN: 3030-3931, Impact factor: 7,241
Volume 8, issue2, Iyul 2025
https://worldlyjournals.com/index.php/Yangiizlanuvchi
worldly knowledge
OAK Index bazalari :
research gate, research bib.
Qo’shimcha index bazalari:
zenodo, open aire. google scholar.
Original article
198
personalized and motivating experience. These applications expose students to diverse English
accents, help identify pronunciation errors, and encourage repeated listening and speaking until
improvement is achieved.
Research and classroom practice have shown that regular engagement with speech recognition
systems enhances phonological awareness—the ability to distinguish between similar sounds and
syllables. Uzbek learners, whose native language differs significantly from English in terms of
phonetics, often find English sounds such as /θ/, /ð/, and final consonant clusters challenging. AI
tools offer targeted exercises and instant correction for these problem areas, which helps to
reinforce correct articulation through repetition and self-monitoring.
Listening skills are also developed more effectively through AI interaction. Since many speech
recognition apps require learners to understand the input accurately before being able to repeat or
respond, students are motivated to pay closer attention to spoken language. The ability to slow
down audio, repeat segments, or receive visual transcripts further supports learners in processing
complex listening tasks. This is particularly helpful for lower-intermediate learners, who may be
overwhelmed by native-speed input in traditional listening recordings.
Moreover, AI speech tools offer an individualized learning path. Unlike classroom settings,
where teachers may not have time to provide detailed feedback to each student, AI systems can
track learner progress, adapt difficulty levels, and highlight specific areas for improvement. This
level of personalization is highly beneficial for Uzbek students, whose language learning
backgrounds and proficiencies vary widely. As a result, learners gain not only technical
improvement but also confidence in their listening and speaking abilities.
Nevertheless, the integration of speech recognition AI into mainstream EFL instruction in
Uzbekistan is not without obstacles. Access to digital devices, internet connectivity, and English-
proficient teachers who can guide students in using these tools remains uneven. In addition, the
accuracy of AI systems may decline when dealing with strong accents or non-native speech
patterns, which can lead to occasional misjudgments in feedback. To mitigate this, training
sessions for teachers and students, along with locally adapted versions of AI tools, are necessary.
Furthermore, motivational factors must be considered. While some learners enjoy the gamified
and interactive nature of AI apps, others may feel frustrated by repetitive tasks or unclear
feedback. Effective integration therefore requires thoughtful instructional design that combines
AI use with human facilitation, peer interaction, and meaningful communicative practice. AI
tools should not replace teachers but serve as powerful supplements that extend practice time,
provide objective data, and support autonomous learning.
In conclusion, speech recognition AI holds significant potential for enhancing listening
comprehension and pronunciation skills among Uzbek EFL learners. By offering immediate
feedback, consistent practice, and personalized learning pathways, these technologies can help
bridge the gap between traditional classroom instruction and real-world language proficiency. To
fully realize this potential, investments in infrastructure, teacher training, and culturally relevant
content are essential. When implemented strategically, AI can empower Uzbek students to
become more confident, accurate, and independent users of English.
References:
ISSN: 3030-3931, Impact factor: 7,241
Volume 8, issue2, Iyul 2025
https://worldlyjournals.com/index.php/Yangiizlanuvchi
worldly knowledge
OAK Index bazalari :
research gate, research bib.
Qo’shimcha index bazalari:
zenodo, open aire. google scholar.
Original article
199
1.
Li, V., & Warschauer, M. (2020).
Emerging technologies and language learning: AI
applications in listening comprehension
.
Language Learning & Technology
, 24(3), 1–15.
– Discusses AI-driven tools in developing listening skills among EFL learners.
2.
Kukulska-Hulme, A. (2020).
Mobile-assisted language learning and speaking:
Practices,
challenges,
and
future
directions
.
ReCALL
,
32(3),
245–264.
– Examines speech technology integration in mobile EFL applications.
3.
Godwin-Jones, R. (2021).
AI and big data in language education: Realities and
expectations
.
Language
Learning
&
Technology
,
25(3),
1–12.
– Provides insight into the use of AI and speech tools in classroom environments.
4.
Derwing, T. M., & Munro, M. J. (2015).
Pronunciation Fundamentals: Evidence-
Based
Perspectives
for
L2
Teaching
and
Research
.
John
Benjamins.
– Establishes the foundations of pronunciation learning relevant to speech AI.
5.
Chiu, T. K. F., & Churchill, D. (2016).
Design of learning environments integrating
mobile
and
AI
technologies
.
Computers
&
Education
,
102,
1–13.
– Presents frameworks for effective tech-based language learning.
6.
Yuldashev, O. & Karimova, D. (2022).
Uzbek learners and the digital shift in language
learning: Challenges and practices
.
Tashkent Journal of Language Pedagogy
, 4(2), 44–57.
– Offers local context on technology-based EFL education in Uzbekistan.
7.
Suvorov, R. (2019).
Automated speech recognition in language learning and assessment:
Review
and
prospects
.
Language
Testing
,
36(4),
523–538.
– Reviews the application of speech recognition in pronunciation feedback.
8.
Chen, J. C. (2016).
The crossroads of English language learners, task-based instruction,
and speech recognition: Advancing autonomy through technology
.
TESOL Journal
, 7(3), 579–
611.
– Highlights autonomy and motivation through AI-supported learning.
9.
Elsa Corp (2023).
AI Speech Technology in Pronunciation Training: Global Impact
Report
.
– Describes global trends and learner outcomes from AI speech tools like Elsa Speak.
UNESCO (2022).
Artificial Intelligence and Education: A guide for policy-makers
. Paris:
UNESCO
Publishing.
– Offers guidelines for ethical and effective integration of AI in education.
