AI-Powered Personalised Learning Platforms for EFL Learning:

Preliminary Results

Raffaella Folgieri

, Marisa Gil

Miriam Bait

and Claudio Lucchiari

Department of Philosophy “Piero Martinetti”, Università degli Studi di Milano, via Festa del Perdono 7, Milan, Italy

Computer Architecture Department, UPC Barcelona TECH, Barcelona, Spain

Keywords: Artificial Intelligence, Learning Platform, EFL, Professional Growth, Education.

Abstract: Artificial intelligence (AI) has been increasingly integrated into the field of education, including personalised

learning platforms. However, concerns have been raised about the potential of AI to replace human teachers

and the impact on student agency and autonomy. In this research, we discuss the development of an AI-

powered platform as a helper, not a substitution, for self-directed personal and professional growth. The

present study investigates the effectiveness of an AI-powered personalised learning platform in enhancing

self-directed learning and personal and professional growth. We also explored the role of human teaching and

the ethical considerations of AI in education. A mixed-methods approach was used, including surveys,

interviews, and qualitative analysis of participant feedback. The participants were randomly assigned to either

an AI group or the traditional learning group. Findings suggest that the AI-powered personalised learning

platform that we used is a promising approach for enhancing self-directed learning and personal and

professional growth. However, it is important to note that these are just preliminary findings, and further

research is needed to confirm our results and to understand the mechanisms by which a specific use of AI in

education may lead to positive effects.

1 INTRODUCTION

Artificial intelligence (AI) has been increasingly

integrated into various fields, including education. In

recent years, personalised learning platforms that use

AI algorithms have emerged as a promising approach

to education. These platforms can provide learners

with personalised feedback, recommendations, and

support based on their learning progress and goals.

The potential of AI-powered personalised

learning platforms to enhance learning outcomes has

been demonstrated in various studies.

Over the past 10 years, research has focused on

the use of technological devices and tools specifically

designed to increase the effectiveness and efficiency

of certain mental processes, such as learning.

The theoretical background is based on the model

of the extended mind (Clark and Chalmers, 1998),

which states that cognitive processes are not confined

https://orcid.org/0000-0002-0589-5275

https://orcid.org/0000-0001-9755-3311

https://orcid.org/0000-0003-0408-2683

https://orcid.org/0000-0002-9452-802X

to the locus of their psychophysical realization, but

extend far beyond the boundaries of the brain,

encompassing not only processes related to bodily

experience but also some external entities, such as

tools, materials of various kinds and technological

devices. In this sense, a key goal of cognitive science

emerges: to study and develop techniques and

technologies that can extend the cognitive capacity,

in a broad sense, of the individual (Roco and

Bainbridge, 2013), seeking to overcome inherent

limitations as well as "cognitive biases" through

technologies that enable the analysis and modulation

of cognition in real time (Heersmink, 2016).

At the same time, the external entities through

which the extended mind manifests and learns can

also be represented by other individuals or social

groups in the context of interpersonal relationships,

thus leading to extending the very concept of the

Folgieri, R., Gil, M., Bait, M. and Lucchiari, C.

AI-Powered Personalised Learning Platforms for EFL Learning: Preliminary Results.

DOI: 10.5220/0012672000003693

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 16th International Conference on Computer Supported Education (CSEDU 2024) - Volume 2, pages 255-261

ISBN: 978-989-758-697-2; ISSN: 2184-5026

255

extended mind (Colombetti and Roberts, 2015)

toward that of collective mind (Tollefsen, 2006).

Building on this theoretical background,

technological tools and creative methods (creative

technology) will be used in neuro-cognitive settings

to implement and validate augmented cognition

methodologies in both individual and dyadic or group

interaction contexts through what is called

collaborative interfaces (Gerhard et al., 2004). It is

self-evident that, in our global world, the role of

English as a lingua franca (ELF) is undeniable.

Online platforms are often used by adult learners

thanks to the flexibility given in time management.

Criticisms stress their lack of human interaction,

but there could also be advantages in adopting this

learning approach, especially if associated with

advanced technology and methodology. In this study,

we aim to explore the thesis that Artificial

Intelligence can improve the level of learning

associated with online platforms as well as avoid any

stress and anxiety possibly connected with native

speakers’ interaction.

However, concerns have been raised about the

potential of AI to replace human teachers and the

impact on student agency and autonomy. There is a

risk that the use of AI in education may lead to a

reduction in the role of human teachers and

consequently a lack of jobs in the sector. Here we

propose an AI-powered personalised learning

platform as a helper, not a substitution, for self-

directed personal and professional growth. However,

the study only aims to compare the learning results

obtained, adopting an AI-powered versus a traditional

approach.

The study aims to answer the following research

questions:

1. What are the advantages of an AI-powered

personalised learning platform in enhancing self-

directed English language compared to traditional

learning methods?

2. What is the potential of AI to provide

personalised feedback and recommendations for self-

directed personal and professional growth?

3. What is the role of human teachers in the

development and implementation of AI platforms?

4. Which ethical considerations about the use of

AI in education should be addressed, and specifically

which ethical principles should inform the

development and implementation of an AI-based

personalised learning platform?

The development of a AI-based personalised

learning system should be considered a support, not a

substitution of traditional approaches aimed to foster

self-directed personal and professional growth.

Indeed, it could provide learners with individual and

customised feedback to support the role of human

teachers.

The present study aims to test the feasibility and

effectiveness of an AI-based learning system. It may

also contribute to the responsible use of AI in

education by addressing ethical considerations and

promoting the collaboration between human teachers

and AI-powered personalised learning platforms.

2 LITERATURE REVIEW

The use of AI in education has been increasing in

recent years. AI has been used in various aspects of

education, including personalised learning platforms

(Simões et al., 2013), intelligent tutoring systems

(Mousavinasab et al., 2021), and adaptive assessment

systems (Osadcha et al., 2022). The potential of AI to

enhance learning outcomes has been demonstrated in

various studies (Alam et al., 2021; Hwang et al.,

2020), with particular regard to the study of English

as a second language (Lotze, 2018; Li, 2020). AI-

powered systems are designed so that they can

provide learners with personalised feedback,

recommendations, and support that can enhance their

learning experiences. These platforms can enhance

self-directed learning by providing learners with the

flexibility to manage their own learning progress and

goals. Self-directed learning is an essential skill for

personal and professional growth in today's rapidly

changing world. Learners need to be able to set their

learning goals, manage their learning progress, and

evaluate their learning outcomes.

However, the use of AI in education should be

approached with caution and ethical considerations

(Hwang et al., 2020; Holmes et al., 2021). There are

concerns about the potential of AI to replace human

teachers (Selwyn, 2019) and the impact on student

agency and autonomy (Hu et al., 2022). Following the

literature, the development and implementation of

AI-powered personalised learning platforms should

prioritise the collaboration between human teachers

and AI algorithms (Humble, 2019) to benefit both

from the human approach and the new technologies.

The potential biases in AI algorithms and the

impact on student privacy and data security should

also be considered in the development and

implementation of AI-powered personalised learning

platforms (Vincent-Lancrin, and R. Van der Vlies,

2020; Razmerita et al., 2022). The responsible use of

AI in education should prioritize ethical

considerations and the promotion of student agency

and autonomy.

CSEDU 2024 - 16th International Conference on Computer Supported Education

256

This work explores the potential of these

platforms through the results of an experiment

conducted in a specifically developed AI-powered

personalised learning platform, namely AIE, for

English language learning. Paragraph 3 illustrates the

experiments; the following paragraph shows the

results, whilst paragraph 5 reports the discussion,

followed by the conclusion.

3 THE EXPERIMENT

3.1 Participants

The participants of this study were adult learners who

are interested in self-directed personal and

professional growth. They have been recruited

through online platforms and social media groups and

were required to have access to a computer or mobile

device with an internet connection.

Table 1: Participants’ characteristics.

Characteristics Group 1 Group 2

Gender Male (n=12),

Female (n=18)

Male (n=15),

Female (n=15)

Age (years) Mean = 25, SD

= 2.5

Mean = 26,

SD = 2.0

Education level High school

diploma (n=5),

Bachelor's degree

(n=20), Master's

degree (n=5)

High school

diploma (n=7),

Bachelor's degree

(n=18),

Master's degree

(

n=5

)

experience with

AI-powered

personalised

learning

platforms

(

self-re

orted

)

Yes (n=10), No

(n=20)

Yes (n=5), No

(n=25)

knowled

Learning English

for the first time

Learning English

for the first time

Native language Italian Italian

All the participants expressed no particular

preference to study using traditional methods or an AI

powered personalised learning platform and they

were all comfortable with technology.

3.2 Platform

An AI-powered personalised learning platform,

namely AIE, for English language learning at the A1

level of the Common European Framework of

Reference for Languages (CEFR), has been designed

to provide an individualised and adaptive learning

experience to each learner, based on the learner's

progress and performance.

AIE starts by assessing the learner's current

English proficiency level using a placement test, and

then create a personalised learning plan based on the

learner's strengths, weaknesses, and learning

preferences.

The platform offers a range of learning materials,

including interactive lessons, quizzes, and exercises,

as well as audio and video resources. In this first

version, AIE tracks the learner's progress over time,

identify areas where the learner needs more practice,

and adjust the level of difficulty of lessons

accordingly. Data used to assess the progress comes

from the results of automated quizzes and tests and

the time spent on each lesson. To capture the

complex, non-linear relationships between the

learner’s performance and their proficiency level, the

platform uses a deep learning approach, specifically a

Convolutional Neural Network (CNN).

The AIE platform has been developed in Python

using TensorFlow libraries. The components are

learner’s profile definition, used to create a

personalised learning plan and adjust the difficulty

level; learning content definition; machine learning

algorithm to analyse the learner’s performance and

adjust the difficulty level; personalised learning plan

definition.

The user interface shows a dashboard that

displays overall progress, completion rates, and quiz

scores. It also provides a button to access the learning

content section, showing the personalised learning

plan. User settings allows the user to customise their

learning experience, changing their learning

preferences, or updating their profile information.

3.3 Procedure

A mixed-methods approach was used, including

surveys, interviews, and qualitative analysis of

participant feedback. The participants were randomly

assigned to either the AI-powered personalised

learning platform group or the traditional learning

methods group.

Table 2: Experimental design.

Grou

Condition

1 AI-powered personalised

learnin

latform

2 Traditional learning methods

The AI group used the platform AIE. The control

group used traditional learning methods, such as

AI-Powered Personalised Learning Platforms for EFL Learning: Preliminary Results

257

lectures and instructional handouts prepared by the

same teachers involved in the design of the

educational material available through the AI-

powered platform. The choice to opt for common

educational material prepared by the same teachers

was made to exclude from the variables the possible

impact of differing qualities of materials in the results

obtained by the students.

The study lasted six weeks. The participants were

required to complete pre- and post-study surveys that

measure their self-directed learning, personal and

professional growth, motivation, and engagement.

The participants have also been required to complete

weekly surveys that measure their learning progress

and feedback on the AI-powered personalised

learning platform or traditional learning methods.

The participants in the AI group have also been

invited to a 30-minute interview at the end of the

study to provide feedback on their experiences with

the platform. The interviews were recorded and

transcribed for qualitative analysis.

3.4 Measures

The study used the following measures:

•

Pre- and post-study surveys that measure

self- directed learning, personal and

professional growth, motivation, and

engagement.

•

Weekly surveys that measure learning

progress and feedback on the AI-powered

personalised learning platform or traditional

learning methods.

•

Interviews with the AI-powered personalised

learning platform group to provide feedback

on their experiences with the platform.

3.5 Data Analysis

The data collected from the surveys have been

analysed using descriptive statistics and inferential

statistics, such as t-tests. The qualitative data

collected from the interviews have been analysed

using thematic analysis. The analysis focused on the

effectiveness of the AI-powered personalised

learning platform in enhancing self-directed learning

and personal and professional growth. The analysis

has also explored the potential of AI to provide

personalised feedback and recommendations and the

role of human teachers in the development and

implementation of AI-powered personalised learning

platforms.

4 RESULTS

The AI-based platform was found to be easy to use

and navigate by the participants. It was specifically

tailored to the learning objectives of the study, which

helped participants stay engaged and motivated

throughout the study.

The platform was designed to personalise the

learning experience for each participant, providing

recommended resources and activities based on their

individual learning needs and preferences.

The platform AIE was found to have a significant

positive impact on self-directed learning and personal

and professional growth and was found to be

significantly more effective than traditional learning

methods as shown in table 3. Considering that in our

teaching experience we could observe that English

courses are generally more difficult for adult

beginners than for young students, the completion

rate was, on the whole, satisfactory, and particularly

high with the support of the platform.

Table 3: Attendance and completion rate.

Group Attendance Rate Completion Rate

1 (AI-powered

personalised

learning platform)

90% 80%

2 (Traditional

learning methods)

85% 70%

Similarly, weekly survey reports confirmed the

positive impact of the AIE platform on learners (see

the following table 4) and the pre- and post-study

survey results (table 5).

Table 4: Weekly survey results.

Week Group 1

(Mean +/- SD)

Group 2

(Mean +/- SD)

P-value

1 3.5 +/- 0.7 3.4 +/- 0.6 0.05

2 3.7 +/- 0.6 3.3 +/- 0.7 0.02

3 4.0 +/- 0.5 3.2 +/- 0.8 0.001

4 4.3 +/- 0.4 3.5 +/- 0.7 0.001

5 4.5 +/- 0.3 3.6 +/- 0.6 0.001

6 4.6 +/- 0.3 3.7 +/- 0.5 0.001

CSEDU 2024 - 16th International Conference on Computer Supported Education

258

Table 5: Pre- and post-study survey results.

Survey

Measure

Group 1

(Mean +/- SD)

Group 2

(Mean +/- SD)

P-value

Self-directed

learning

Pre-test: 3.2

+/-

0.9, Post-

test: 4.5

+/- 0.7

Pre-test: 3.1

+/-

0.8, Post-

test: 3.2

+/- 0.9

0.001

Personal and

professional

growth

Pre-test: 2.8

+/-

0.6, Post-

test: 4.1

+/- 0.5

Pre-test: 2.9

+/-

0.7, Post-

test: 3.0

+/- 0.8

0.003

Motivation Pre-test: 3.5

+/-

0.8, Post-

test: 4.6

+/- 0.6

Pre-test: 3.4

+/-

0.7, Post-

test: 3.5

+/- 0.8

0.001

Engagement Pre-test: 3.0

+/-

0.7, Post-

test: 4.2

+/- 0.5

Pre-test: 3.1

+/-

0.6, Post-

test: 3.3

+/- 0.7

0.001

To answer the question if there is a significant

difference from self-learning through the AI-powered

platform and the traditional learning approach, data

were analysed through a two-tails t-test. The data

shows that there is a significant difference in the mean

of the two groups for all four survey measures. In

other words, there is a statistically significant

difference between the mean self-directed learning

scores for the two groups, the mean personal and

professional growth scores for the two groups, the

mean motivation scores for the two groups, and the

mean engagement scores for the two groups.

Based on the p-values reported in tables 4 and 5,

the experimental questions for which the obtained p-

values are as low as 0.001 are:

Is the AI-powered personalised learning

platform more effective than traditional

learning methods in enhancing self-directed

English language skills and personal and

professional growth?

Is the AI-powered personalised learning

platform more effective than traditional

learning methods in enhancing the motivation

and engagement of learners?

These findings suggest that the AI-powered

personalised learning platform that we used in the

present study is a promising approach for enhancing

self-directed learning and personal and professional

growth. However, it is important to note that these are

just preliminary findings, and further research is

needed to confirm these results and to understand the

mechanisms by which an AI-based platform works

and obtain the observed results.

Table 6: Attendance and completion rate.

Group

Pre-Test Score

(Mean +/- SD)

Post-Test Score

(Mean +/- SD)

P-value

75 +/- 5 85 +/- 5 0.001

73 +/- 6 75 +/- 7 0.05

Following Table 6, the cognitive test results

showed that participants in the platform group had

significantly higher scores at post-test compared to.

The platform AIE allowed participants to learn at

their own pace and provided personalised feedback

and recommendations, which helped them feel more

in control of their learning and supported their

personal and professional growth. The participants

expressed their positive opinion during the interview.

The AI-powered personalised learning platform was

found to have great potential for providing

personalised feedback and recommendations.

Participants reported finding the feedback and

recommendations to be useful and relevant to their

learning objectives, as reported in table 7 and this

personalised approach to learning was a key factor in

the platform's effectiveness.

Table 7: Interview feedback.

Theme Group 1 Group 2

Ease of use 90% positive 70%

positive

Relevance to

learning objectives

95% positive 80%

positive

Engagement 85% positive 60%

positive

5 DISCUSSION

5.1 Summary of Findings

The study found that the AI-based system used was

significantly more effective in promoting self-

directed learning and personal and professional

growth than traditional learning methods. The

platform provided personalised recommendations for

resources and activities based on each participant's

individual learning needs, which helped them feel

more in control of their learning and supported their

AI-Powered Personalised Learning Platforms for EFL Learning: Preliminary Results

259

personal and professional growth. The platform was

also found to be highly engaging and motivated

participants to stay in the learning process.

5.2 Implications of the Study

The study has important implications for the future of

education. The use of AI-powered personalised

learning platforms has the potential to revolutionise

education by providing tailored feedback and

recommendations to support individual learning

needs and preferences. This personalised approach to

learning has the potential to significantly improve the

effectiveness of education and it may also support

personal and professional growth of learners. The

findings of the study suggest that incorporating

tailored AI applications into educational settings

could be a valuable tool for enhancing learning

outcomes, also complementing traditional

approaches.

5.3 Limitations and Future Directions

The present study has some limitations that should be

taken into consideration. First, the study was

conducted with a relatively small sample size and in

a specific educational setting, which limits the

generalizability of the findings. Additionally, the

study only measured short-term outcomes, and it is

unclear how the effects of the platform will persist

over time. Finally, in a further version, also other

algorithms could be included in the platform, so that

a more general approach to the evaluation of the

results could be provided, for example considering an

ensemble of AI methods.

Future research should focus on addressing these

limitations and exploring the long-term effects of AI-

powered personalised learning platforms on learning

outcomes. Additionally, it will be important to

investigate ways to ensure that the algorithms used by

such platforms are fair and unbiased.

Furthermore, future studies could include the

recording of bio-signals (e.g. EEG and electro-dermal

activity through the use of wearable devices) which

will allow observing and quantifying neurocognitive

correlates of the educational process elicited by the

AI platform use. The collected data would allow a

deeper understanding of the cognitive process

implied by this particular learning context, also

permitting a further development of the technological

architecture of the platform so to adapt to cognitive

and psychological features of learners.

6 CONCLUSION

In conclusion, the study found that the AI-powered

personalised learning platform specifically

implemented for this study was significantly more

effective in promoting self-directed learning and

personal and professional growth compared to

traditional learning methods. The platform provided

personalised recommendations for resources and

activities based on each participant's individual

learning needs, which helped them feel more in

control of their learning and supported their personal

and professional growth. The study has important

implications for the future of education, and further

research is needed to address the limitations of the

study and explore the potential of AI-powered

personalised learning platforms to revolutionise

education. AI algorithms cannot replace the role of

the human teachers in guiding the development of

critical thinking and creativity. What's more, the

training provided by AI needs human guidance to

allow the growth of each student's personality.

Although AIE showed its potential in education,

during the experiment the human teacher is essential

to provide feedback and guide the results obtained so

that the platform can suitably adapt each student's

learning and direction of development. This fact was

clear from the interviews conducted weekly and at the

end of the experiment. This collaboration between AI

training and human teaching is a unique opportunity

for teachers to help the students reach a flourishment

adapted and specific for each one. This observation

reinforces the idea of human-platform collaboration

and the possibilities also open to more creative and

person-oriented teacher tasks.

ACKNOWLEDGEMENTS

This research was funded by the Department of

Philosophy “Piero Martinetti” of the University of

Milan under the Project “Departments of Excellence

2023-2027” awarded by the Ministry of University

and Research (MUR).

Marisa Gil's work has been supported by the

Spanish Ministry of Education (PID2019-107255GB-

C22) and the Generalitat de Catalunya (2021-SGR-

01007).

CSEDU 2024 - 16th International Conference on Computer Supported Education

260

REFERENCES

Alam, A. (2021, November). Possibilities and

apprehensions in the landscape of artificial intelligence

in education. In 2021 International Conference on

Computational Intelligence and Computing

Applications (ICCICA) (pp. 1-8). IEEE.

Clark, A., & Chalmers, D. (1998). The extended mind.

analysis, 58(1), 7-19.

Colombetti, G., & Roberts, T. (2015). Extending the

extended mind: the case for extended affectivity.

Philosophical Studies, 172(5), 1243-1263.

Gerhard, M., Moore, D., & Hobbs, D. (2004). Embodiment

and copresence in collaborative interfaces. International

Journal of Human-Computer Studies, 61(4), 453-480.

Heersmink, R. (2016). The internet, cognitive

enhancement, and the values of cognition. Minds and

Machines, 26(4), 389-407.

Holmes, W., Porayska-Pomsta, K., Holstein, K.,

Sutherland, E., Baker, T., Shum, S. B., ... & Koedinger,

K. R. (2021). Ethics of AI in education: Towards a

community-wide framework. International Journal of

Artificial Intelligence in Education, 1-23.

Hu, X., Liu, Y., Huang, J., & Mu, S. (2022). The Effects of

Different Patterns of Group Collaborative Learning on

Fourth-Grade Students’ Creative Thinking in a Digital

Artificial Intelligence Course. Sustainability, 14(19),

12674.

Humble, N., & Mozelius, P. (2019, October). Teacher-

supported AI or AI-supported teachers. In European

Conference on the Impact of Artificial Intelligence and

Robotics (ECIAIR 2019) (pp. 157-164).

Hwang, G. J., Xie, H., Wah, B. W., & Gašević, D. (2020).

Vision, challenges, roles and research issues of

Artificial Intelligence in Education. Computers and

Education: Artificial Intelligence, 1, 100001.

Li, R. (2020, November). Using artificial intelligence in

learning English as a foreign language: an examination

of IELTS LIULISHUO as an online platform. In

Conference Proceedings. Innovation in Language

Learning 2020.

Lotze, N. (2018). Goodbye to classroom teaching?

Artificial intelligence in language learning.

https://www.goethe.de/en/spr/eng/gls/21290629.html

(accessed on 27 April 2023).

Mousavinasab, E., Zarifsanaiey, N., R. Niakan Kalhori, S.,

Rakhshan, M., Keikha, L., & Ghazi Saeedi, M. (2021).

Intelligent tutoring systems: a systematic review of

characteristics, applications, and evaluation methods.

Interactive Learning Environments, 29(1), 142-163.

Osadcha, K. P., Osadchyi, V. V., Kruglyk, V. S., & Spirin,

O. M. (2022). Analysis and summarization of the

experience of developing adaptive learning systems in

higher education. In AET 2020: Proceedings of the 1st

Symposium on Advances in Educational Technology

(No. 2, pp. 208-215). SCITEPRESS– Science and

Technology Publications.

Razmerita, L., Brun, A., & Nabeth, T. (2022).

Collaboration in the machine age: Trustworthy human-

AI collaboration. In Advances in Selected Artificial

Intelligence Areas: World Outstanding Women in

Artificial Intelligence (pp. 333-356). Cham: Springer

International Publishing.

Roco, M. C., & Bainbridge, W. S. (Eds.). (2013).

Converging technologies for improving human

performance: Nanotechnology, biotechnology,

information technology and cognitive science. Springer

Science & Business Media.

Selwyn, N. (2019). Should robots replace teachers?: AI

and the future of education. John Wiley & Sons.

Simões, T. M., Rodrigues, J. J., & de la Torre, I. (2013).

Personal Learning Environment Box (PLEBOX): A

new approach to E‐learning platforms. Computer

Applications in Engineering Education, 21(S1), E100-

E109.

Speer, P. W., Peterson, N. A., Christens, B. D., & Reid, R.

J. (2019). Youth cognitive empowerment:

Development and evaluation of an instrument.

American journal of community psychology, 64(3-4),

528-540.

Tollefsen, D. P. (2006). From extended mind to collective

mind. Cognitive systems research, 7(2-3), 140-150.

Vincent-Lancrin, S., & Van der Vlies, R. (2020).

Trustworthy artificial intelligence (AI) in education:

Promises and challenges.

AI-Powered Personalised Learning Platforms for EFL Learning: Preliminary Results

261