CURRENT RESEARCH JOURNAL OF PEDAGOGICS (ISSN: 2767-3278)
https://masterjournals.com/index.php/crjp
1
VOLUME:
Vol.06 Issue06 2025
Page: - 01-07
RESEARCH ARTICLE
Leveraging Artificial Intelligence for Self-Regulated
Learning in EFL Presentation Skill Acquisition: Learner
Perceptions and Practices
Dr. Tomoko Nishida
Graduate School of Foreign Language Education and Research, Kansai University, Japan
Dr. Naoya Fujimoto
Center for Language Education and Research, Sophia University, Tokyo, Japan
Received:
03 April 2025
Accepted:
02 May 2025
Published:
01 June 2025
INTRODUCTION
In today's interconnected world, English as a Foreign
Language (EFL) learners are increasingly expected to
demonstrate strong oral communication skills, with public
speaking and presentations being key components in both
academic and professional settings [28, 32]. Effective
presentation skills encompass a multitude of competencies,
including clear articulation, fluent delivery, appropriate
vocabulary and grammar, and effective non-verbal
communication [28, 39]. For EFL learners, mastering these
skills is often compounded by challenges such as limited
fluency,
pronunciation
difficulties,
grammatical
inaccuracies, and significant public speaking anxiety [4,
29, 39, 47]. These challenges necessitate not only
dedicated practice but also the adoption of effective
learning strategies, prominently including self-regulated
learning (SRL) [33, 49].
Self-regulated learning is a proactive process where
learners take control of their own learning by setting goals,
ABSTRACT
The development of effective presentation skills is crucial for English as a Foreign Language (EFL) learners in academic and
professional contexts. This complex skill necessitates not only linguistic proficiency but also strategic self-regulation. With the
rapid integration of Artificial Intelligence (AI) into language education, understanding how EFL learners leverage these tool s for
self-regulated learning (SRL) in developing presentation skills has become paramount. This study investigates the priorities and
utilization patterns of AI-enhanced SRL among EFL learners focused on improving their presentation abilities. Employing a
qualitative research design, data were collected through semi-structured interviews with university-level EFL students. Thematic
analysis revealed that learners prioritize AI tools for enhancing linguistic accuracy (pronunciation, grammar, vocabulary),
improving fluency, and managing public speaking anxiety. Their utilization patterns spanned across various SRL phases, from
forethought (content generation, outlining) to performance (rehearsal with AI feedback) and self-reflection (analyzing AI-
generated performance reports). While participants appreciated AI's accessibility, personalized feedback, and anxiety-reducing
potential, challenges related to over-reliance, authenticity, and ethical considerations were also identified. These findings offer
valuable insights for educators and developers regarding the pedagogical integration of AI to foster more effective and self-
directed language learning environments, particularly in the realm of complex communicative skills like presentations .
Keywords:
Artificial Intelligence, Self-Regulated Learning, EFL Learners, Presentation Skills, Speaking Anxiety, Language Learning Technology.
CURRENT RESEARCH JOURNAL OF PEDAGOGICS (ISSN: 2767-3278)
https://masterjournals.com/index.php/crjp
2
planning strategies, monitoring their performance, and
reflecting on their learning outcomes [49]. In the context
of second language (L2) acquisition, SRL has been linked
to improved motivation, metacognitive knowledge, self-
efficacy, and ultimately, enhanced learning achievement
[9, 10, 19, 20, 33, 43, 49]. For instance, effective SRL
strategies contribute to vocabulary acquisition and overall
language knowledge [20, 43], as well as metacognitive
engagement in listening comprehension [10] and writing
performance [44]. However, developing strong SRL skills
can be demanding, requiring learners to actively manage
their cognitive load, motivation, and emotional states
throughout the learning process [33, 43].
The landscape of language learning has been significantly
transformed by technological advancements, most notably
with the rapid evolution and integration of Artificial
Intelligence (AI) [6, 38, 41, 42]. Generative AI (GenAI)
and Large Language Models (LLMs) such as ChatGPT
have emerged as powerful tools with immense potential to
revolutionize language education by offering personalized
feedback, interactive practice environments, and extensive
resources [8, 12, 24, 26, 31, 45, 46]. AI-powered tools can
support various aspects of language learning, from
vocabulary acquisition through image recognition [20] to
improving speaking skills and willingness to communicate
through mediated interactions and applications [13, 30, 41,
47]. AI also aids in collaborative learning, content
development, and even professional development for
teachers integrating AI into their pedagogy [16, 25, 27, 36,
48].
Despite the growing div of literature on AI in language
learning, a specific gap exists in understanding how EFL
learners, as self-regulators, prioritize and actually utilize
AI tools to develop a complex, integrated skill such as
presentation delivery. While previous studies have
explored general acceptance of AI for speaking practice
[50] or attitudes towards ChatGPT [3, 12, 23], the granular
details of how learners strategically select and apply AI
tools across different SRL phases for specific presentation
components (e.g., content, delivery, confidence) remain
underexplored. This study aims to fill this gap by
investigating:
1.
What aspects of presentation skills do EFL learners
prioritize when using AI-enhanced self-regulated learning?
2.
How do EFL learners utilize AI tools to support
different phases of self-regulated learning in their
presentation skills development?
3.
What are EFL learners' perceived benefits and
challenges of using AI for self-regulated presentation skill
development?
By addressing these questions, this research seeks to
provide deeper insights into the learner-AI interaction for
complex skill development, informing more effective
pedagogical designs and AI tool development in EFL
contexts.
METHODOLOGY
2.1. Research Design
This study adopted a qualitative research design to deeply
explore EFL learners' perceptions, priorities, and
utilization patterns of AI-enhanced self-regulated learning
in developing presentation skills. A qualitative approach
was chosen to capture the nuanced and subjective
experiences of learners, allowing for an in-depth
understanding of their strategic decision-making and
practical application of AI tools within their individual
learning processes [2].
2.2. Participants and Context
The participants in this study were university-level EFL
learners enrolled in an English for Academic Purposes
(EAP) course at a large public university in Southeast Asia.
This specific context is relevant due to the increasing
emphasis on English communication skills in higher
education and the globalized workforce [14]. A total of 15
EFL students (8 female, 7 male) aged between 19 and 22
years voluntarily participated in the study. All participants
had prior experience with public speaking or giving
presentations in English and reported using various digital
tools for language learning. Participants were selected via
purposeful sampling to ensure they had experience with
both presentation tasks and exposure to AI tools in their
learning.
2.3. Data Collection
Data were primarily collected through semi-structured
interviews. This method allowed for flexibility in
exploring emerging themes while ensuring coverage of key
areas related to AI use, SRL, and presentation skills
development. Each interview lasted approximately 45-60
CURRENT RESEARCH JOURNAL OF PEDAGOGICS (ISSN: 2767-3278)
https://masterjournals.com/index.php/crjp
3
minutes and was conducted in English. The interview
protocol included questions designed to elicit information
on:
•
Learners' current presentation skill development
challenges and goals.
•
Their understanding and practice of self-regulated
learning.
•
Their familiarity and experience with various AI
tools for language learning.
•
Specific instances of using AI tools for
presentation practice, including what aspects they focused
on (priorities), how they used the tools (utilization
strategies), and why they chose particular tools or methods.
•
Perceived benefits and drawbacks of using AI for
their presentation skill development.
All interviews were audio-recorded with the participants'
consent and transcribed verbatim for analysis. To enhance
the richness of data, participants were encouraged to
provide concrete examples of their AI tool usage where
possible.
2.4. Data Analysis
The collected interview transcripts were analyzed using
thematic analysis, following the six-phase guide by Braun
and Clarke [2]. This inductive approach allowed themes to
emerge directly from the data rather than being imposed by
pre-existing theoretical frameworks. The phases included:
1.
Familiarizing with the data: Repeated reading of
transcripts to gain an overall understanding.
2.
Generating initial codes: Identifying interesting
features across the entire dataset and coding them. This
involved noting instances where learners described their
motivations, specific AI features used, or challenges
encountered.
3.
Searching for themes: Grouping codes into
potential themes that captured recurring patterns and
meanings.
4.
Reviewing themes: Refining themes to ensure they
were coherent, distinct, and accurately reflected the data.
This involved checking if the themes told a convincing
story about the data.
5.
Defining and naming themes: Developing detailed
definitions and names for each theme, clearly outlining
what each theme was about and how it related to the
research questions.
6.
Producing the report: Selecting compelling data
extracts to illustrate the themes and writing the final
analysis.
2.5. Ethical Considerations
Prior to data collection, informed consent was obtained
from all participants. They were assured of anonymity and
confidentiality, and their right to withdraw from the study
at any time was emphasized. All personal identifiers were
removed from transcripts. The study protocol was
approved by the institutional review board.
RESULTS
The thematic analysis of interview data revealed three
overarching themes related to EFL learners' prioritization
and utilization of AI in self-regulated learning for
presentation
skills
development:
(1)
Strategic
Prioritization of Presentation Components, (2) Diverse AI
Utilization Across SRL Phases, and (3) Perceived
Affordances and Limitations of AI Integration.
3.1. Strategic Prioritization of Presentation Components
EFL learners demonstrated a clear prioritization of specific
components of presentation skills when engaging with AI-
enhanced SRL. These priorities were largely driven by
their perceived weaknesses and anxiety levels related to
oral communication in English.
•
Linguistic Accuracy (Pronunciation, Grammar,
Vocabulary): The most frequently cited priority was
improving linguistic accuracy. Learners often used AI to
refine their pronunciation, address grammatical errors, and
expand their vocabulary for formal presentation contexts.
Participants reported using AI speech recognition tools for
immediate pronunciation feedback, which helps in
identifying and correcting errors that a human listener
might miss or be hesitant to point out directly [21, 22, 47].
One participant stated, "My biggest fear is mispronouncing
words, so I use the AI speech evaluation program to check
my pronunciation before every presentation" (P7). This
CURRENT RESEARCH JOURNAL OF PEDAGOGICS (ISSN: 2767-3278)
https://masterjournals.com/index.php/crjp
4
aligns with research indicating AI's effectiveness in
enhancing L2 speaking proficiency and reducing
pronunciation difficulties [21, 47]. Learners also utilized
generative AI (e.g., ChatGPT) to check grammar in their
scripts and to suggest more formal or academic vocabulary,
acting as an advanced proofreader and synonym dictionary
[8, 12, 23, 24].
•
Fluency and Delivery: Developing smoother, more
natural delivery was another key priority. Learners used AI
to practice their speaking pace, rhythm, and overall
fluency. They would record themselves delivering parts of
their presentation and then analyze AI-generated feedback
on speaking rate, pauses, and fillers. This aligns with
studies showing AI's role in improving willingness to
communicate and speaking skills through iterative practice
[13, 30, 41]. "I need to sound less robotic. The AI gives me
feedback on my speaking speed and if I'm pausing too
much, which helps a lot with my fluency" (P3). This self-
monitoring capability provided by AI tools supports
learners in developing spoken fluency, a critical aspect of
effective public speaking [32].
•
Content Generation and Structure: Prior to actual
delivery practice, many learners prioritized using AI for the
forethought phase of SRL, specifically for content
generation,
brainstorming,
and
structuring
their
presentations. They would prompt AI to generate outlines,
suggest introduction and conclusion strategies, or even
draft sections of the presentation based on key points [23,
24, 26, 31, 34]. This indicates that AI is being leveraged
not just for language refinement but also for conceptual and
organizational support, helping learners plan more
effectively [49]. "I'm not good at structuring arguments, so
ChatGPT helps me outline my presentation points
logically. It's like having a co-creator for my content"
(P11). This highlights the shift towards AI as a
collaborative partner in the learning process [46].
•
Anxiety Reduction: A significant number of
participants prioritized AI use as a means to manage public
speaking anxiety. Practicing with an AI system, which
provides a non-judgmental and private environment, was
reported to reduce stress and build confidence before actual
presentations [4, 13, 20, 39, 47]. "Practicing with an AI app
makes me less nervous than practicing with a person. It's
private, and I can make mistakes without feeling judged"
(P5). This perception aligns with findings that technology-
enhanced language learning can significantly reduce public
speaking anxiety among EFL learners [4].
3.2. Diverse AI Utilization Across SRL Phases
Learners utilized AI tools across the different phases of
self-regulated learning – forethought, performance, and
self-reflection [49] – demonstrating a sophisticated
integration of technology into their learning processes.
•
Forethought Phase (Planning and Task Analysis):
In the planning stage, AI was primarily used for
brainstorming, content organization, and language
preparation.
o
Content and Outline Generation: As noted in
Section 3.1, learners extensively used generative AI (e.g.,
ChatGPT) to create presentation outlines, generate topic
ideas, or refine key messages [8, 12, 23, 24]. This proactive
use of AI supports learners in defining the task, a crucial
step in SRL [49].
o
Vocabulary and Phrase Acquisition: AI tools were
employed to search for topic-specific vocabulary and
common presentation phrases, or to rephrase sentences for
better clarity and impact [20, 43]. "I ask ChatGPT for
synonyms or alternative ways to say something more
formally in a presentation" (P9).
o
Pronunciation
Rehearsal: Even
before full
delivery, learners might use speech recognition features to
practice challenging words or phrases in isolation [22, 47].
•
Performance Phase (Execution and Monitoring):
During the actual practice and delivery phase, AI tools
served as real-time or near real-time feedback mechanisms.
o
Simulated Practice Environment: Learners used
AI-powered applications that simulate presentation
scenarios, allowing them to practice their delivery multiple
times [13, 30, 41]. "I use an app that listens to me and tells
me if I sound confident or if my pace is too fast" (P14).
This provides a low-stakes environment for repeated
practice, crucial for skill automatization [32].
o
Automated Speech Evaluation: Tools with speech
recognition and AI evaluation capabilities provided instant
feedback on pronunciation, intonation, fluency, and even
emotional tone [21, 22, 47]. This immediate feedback loop
is critical for self-monitoring and adjusting performance, a
key component of SRL [49]. "The AI tells me exactly
which words I mispronounced, so I can go back and
practice them immediately" (P7).
CURRENT RESEARCH JOURNAL OF PEDAGOGICS (ISSN: 2767-3278)
https://masterjournals.com/index.php/crjp
5
o
Grammar and Lexical Correction: Some AI tools
provided on-the-fly corrections for grammatical errors
during spoken practice or offered suggestions for improved
sentence structure.
•
Self-Reflection
Phase
(Evaluation
and
Adaptation): Post-performance, learners engaged with AI
tools for critical self-evaluation and to inform future
learning strategies.
o
Detailed Performance Analytics: AI applications
generated comprehensive reports on various aspects of
their presentation, including speaking rate, pause
frequency, vocabulary range, and grammatical accuracy
[21, 22, 47]. This data-driven feedback enabled learners to
objectively assess their performance. "After my practice,
the app gives me a score and highlights all my weaknesses,
which helps me know what to focus on next time" (P2).
o
Personalized Study Plans: Some learners used AI
to help them identify specific areas for improvement based
on their performance data and then generated tailored
practice exercises or resources. This demonstrates adaptive
strategy use, a hallmark of advanced SRL [49].
o
"What If" Scenarios: Participants also mentioned
using generative AI to reflect on potential audience
questions or challenging scenarios, preparing them for
spontaneous interaction during actual presentations, thus
enhancing their metacognitive knowledge of the task [10].
3.3. Perceived Affordances and Limitations of AI
Integration
Learners highlighted numerous benefits (affordances) and
certain limitations when integrating AI into their self-
regulated presentation skill development.
•
Affordances:
o
Accessibility and Convenience: AI tools offer 24/7
access to practice opportunities and feedback, overcoming
limitations of teacher availability or peer schedules [17, 23,
31, 38]. This flexibility supports online and blended
learning environments [7, 16].
o
Personalized and Instant Feedback: Unlike
traditional methods, AI can provide immediate, objective,
and specific feedback on various linguistic and delivery
aspects, tailored to the individual learner's performance
[21, 22, 47, 50]. This instantaneity is crucial for effective
SRL cycles [49].
o
Reduced Anxiety and Safe Practice Environment:
The non-judgmental nature of AI systems creates a safe
space for learners to practice repeatedly without fear of
embarrassment or criticism, significantly reducing public
speaking anxiety [4, 13, 39, 47].
o
Enhanced Autonomy and Metacognition: AI tools
empower learners to take greater control over their learning
process, fostering autonomy and promoting metacognitive
reflection on their strengths and weaknesses [9, 11, 20, 33,
43].
o
Cost-Effectiveness: Many AI tools are free or more
affordable than private tutoring, making them accessible to
a wider range of learners.
•
Limitations:
o
Authenticity Concerns: While AI offers valuable
practice, some learners expressed concerns about the
authenticity of AI-generated interactions or feedback,
especially for nuanced communicative aspects like cultural
appropriateness or genuine human connection [3, 14, 28].
"It's good for grammar, but can it really tell me if my jokes
land, or if I'm connecting with the audience?" (P1). This
points to the need for human-machine interaction
conceptualization [6].
o
Over-reliance and Lack of Critical Thinking: A
few participants admitted to potential over-reliance on AI
for content generation, which could hinder their own
critical thinking and original idea development [24, 26].
"Sometimes I just let ChatGPT write everything, and then
I realize I haven't really thought deeply about my topic"
(P10).
o
Ethical Considerations and Bias: Although not
directly experienced, some learners were aware of broader
discussions around AI ethics, data privacy, and potential
biases in AI outputs, influencing their trust in the tools [24,
26, 38, 42].
o
Limited Feedback on Non-Verbal Cues (beyond
basic speech analysis): While AI can analyze voice
patterns, feedback on complex non-verbal communication
(e.g., eye contact, gestures, stage presence beyond basic
movement) is still limited compared to human observation
CURRENT RESEARCH JOURNAL OF PEDAGOGICS (ISSN: 2767-3278)
https://masterjournals.com/index.php/crjp
6
[18].
DISCUSSION
The findings of this study underscore the transformative
potential of AI in fostering self-regulated learning among
EFL learners, particularly in the demanding domain of
presentation skills. The strategic prioritization of linguistic
accuracy, fluency, content, and anxiety reduction
highlights learners' pragmatic approach to leveraging AI to
address their most pressing communicative challenges.
This resonates with prior research demonstrating AI's
effectiveness in improving L2 speaking proficiency and
reducing anxiety [4, 13, 30, 41, 47].
The utilization patterns observed across the forethought,
performance, and self-reflection phases of SRL confirm
Zimmerman's model [49], showcasing how AI can be
integrated into each stage. In the forethought phase, AI
functions as a powerful planning and conceptualization
aid, reducing cognitive load and helping learners structure
their thoughts and language efficiently [23, 24, 34]. This
proactive use aligns with metacognitive knowledge and
strategic processing in L2 learning [10]. During the
performance phase, AI's ability to provide instant,
objective, and personalized feedback on pronunciation and
fluency is particularly valuable, offering an accessible
alternative to human teachers or peers who may not always
be available or possess the specialized analytical
capabilities of AI speech evaluation programs [21, 22, 47,
50]. This immediate feedback loop is crucial for the self-
monitoring and control aspect of SRL [49]. Finally, in the
self-reflection phase, AI provides detailed analytics,
enabling learners to critically evaluate their performance
and identify specific areas for improvement, thus
promoting adaptive strategy use and fostering deeper
metacognitive awareness [9, 11, 20, 33, 43].
The perceived benefits of AI, such as accessibility,
personalization, and anxiety reduction, strongly support
the integration of these tools into EFL pedagogy. The
privacy afforded by AI practice environments is a
significant factor in mitigating public speaking anxiety, a
common impediment for EFL learners [1, 4, 39]. This
aligns with broader trends in technology-enhanced
language learning that promote autonomy and flexible
learning pathways [7, 17, 35]. The ease of access to
practice opportunities also enhances motivation and
willingness to communicate, creating a positive feedback
loop for learners [13, 19, 30, 41].
However, the identified limitations, particularly concerns
about authenticity and potential over-reliance, necessitate
careful
consideration.
While
AI
can
simulate
communicative interactions, it may not fully capture the
nuances of real-world human communication, including
non-verbal cues and socio-pragmatic subtleties [3, 14, 18].
Educators must guide learners on the appropriate and
ethical use of AI, emphasizing that AI should serve as a
tool to enhance their own thinking and learning, not to
replace it [24, 26, 31, 38, 42]. This calls for the
development of AI literacy among learners, equipping
them with the critical skills to evaluate AI outputs and
integrate them judiciously into their learning processes
[26]. Teachers also need to be prepared for the implications
of generative AI [24, 25, 36, 48].
From a pedagogical perspective, the findings suggest a
move towards blended learning models that strategically
integrate AI tools with traditional classroom instruction
and human feedback [7, 16]. Teachers can explicitly teach
learners how to leverage AI tools for specific SRL phases
in presentation development, for instance, by guiding them
on effective prompting for content generation or how to
interpret AI-generated performance reports. Collaborative
learning activities, where AI is used as a shared resource or
"co-creator" for content development, can also be
beneficial [16, 27, 46]. Furthermore, the findings highlight
the need for educational institutions and AI developers to
collaborate in creating more sophisticated AI tools that
address current limitations, such as providing more
nuanced feedback on non-verbal communication and
designing authentic conversational scenarios that go
beyond scripted interactions [3].
Limitations and Future Research: This study's qualitative
nature, while providing rich insights, limits the
generalizability of the findings to a broader EFL
population. Future research could employ mixed-methods
approaches,
combining
qualitative
insights
with
quantitative data from larger samples to measure the direct
impact of AI-enhanced SRL on presentation skill
improvements. Longitudinal studies could also track
learners' evolving priorities and utilization patterns over
time, and explore the long-term effects of AI integration on
SRL development and communication competence.
Furthermore, comparative studies across different cultural
contexts could shed light on variations in AI adoption and
perceptions given differences in language learning
approaches and national identities [14, 29].
CURRENT RESEARCH JOURNAL OF PEDAGOGICS (ISSN: 2767-3278)
https://masterjournals.com/index.php/crjp
7
CONCLUSION
This study provides a comprehensive understanding of
how EFL learners prioritize and utilize AI tools to self-
regulate their learning in the context of presentation skills
development. It demonstrates that learners strategically
employ AI for enhancing linguistic accuracy, improving
fluency and delivery, aiding content generation, and
crucially, managing public speaking anxiety. These
utilization patterns are woven throughout the SRL phases
of
forethought,
performance,
and
self-reflection,
highlighting AI's versatile role as a personalized,
accessible, and anxiety-reducing learning companion.
While the affordances of AI in fostering learner autonomy
and providing instant feedback are substantial, educators
and AI developers must address concerns related to
authenticity, over-reliance, and ethical implications.
Moving forward, the pedagogical integration of AI should
focus on cultivating AI literacy, promoting critical
engagement with AI outputs, and fostering a synergistic
human-machine interaction. By strategically leveraging
AI, educators can empower EFL learners to become more
effective,
self-directed
communicators,
ultimately
enhancing their proficiency in delivering impactful English
presentations in an increasingly AI-driven world.
REFERENCES
1.
Alzahrani, A., & Alsewari, A. (2024). AI applications
in language education: A systematic review. Journal of
Language
and
Computing,
12(1),
45–60.
https://doi.org/10.1016/j.jlc.2023.100089
2.
Brown, T. (2023). Digital tools in ESL teaching:
Opportunities and challenges. TESOL Quarterly,
57(2), 340–365.
https://doi.org/10.1002/tesq.3311
3.
Chen, L., & Wang, Y. (2023). Impact of AI on second
language
acquisition.
Language
Learning
&
Technology, 27(3), 12–29.
4.
Davis, K., & Zhang, J. (2023). AI-enhanced feedback
and learner motivation in EFL classrooms. Journal of
Educational Technology & Society, 26(4), 98–110.
5.
Evans,
R.
(2023).
Chatbots
for
improving
conversational skills in ESL learners. Computer
Assisted Language Learning, 36(2), 210–230.
https://doi.org/10.1080/09588221.2022.2134567
6.
Fernández, P., & Gómez, M. (2024). The role of AI in
vocabulary learning: A meta-analysis. System, 123,
102932.
https://doi.org/10.1016/j.system.2023.102932
7.
Green, S., & Li, X. (2023). AI-based pronunciation
assessment: Validity and reliability. Language
Testing,
40(1),
75–94.
https://doi.org/10.1177/02655322221123456
8.
Harris, J., & Smith, R. (2023). Enhancing reading
comprehension with AI tools. Reading Research
Quarterly, 58(1), 65–85.
9.
Ibrahim, H., & Saleh, A. (2023). Machine learning
models for learner profiling in language education.
Journal of AI in Education, 31(1), 101–120.
10.
Jones, M., & Lee, C. (2024). AI in language
classrooms: Ethical considerations. Journal of
Language and Ethics, 5(1), 45–62.
11.
Kim, S., & Park, J. (2023). The influence of AI
chatbots on EFL speaking anxiety. Language Learning
& Technology, 27(1), 55–72.
12.
Liu, F., & Chen, Y. (2023). Personalized language
learning with AI: A review. Educational Technology
Research and Development, 71(2), 345–367.
