Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

Implementation of a Personalized
Adaptive Mobile Learning System
P.E. Akinwole1; O. K. Boyinbode2; M.T. Kinga 3; P.K. Olotu4
Department of Information Technology, Federal University of Technology, Akure, Nigeria

Abstract:- A notable obstacle in the field of education is Personalization in learning encompasses adapting
the restricted flexibility of traditional teaching content, pace, instructional strategies, and assessment
approaches. These approaches frequently take a methods to align with learners' unique characteristics and
consistent stance, disregarding the wide range of needs. This individualized approach enhances learner
learning preferences that pupils possess. This leads to a engagement, motivation, and learning outcomes
decrease in student motivation and engagement, which significantly.
in turn produces below-average learning outcomes. This
research focuses on creating an adaptive learning system The incorporation of learning style models, such as the
that classifies learners using the Felder-Silverman model Felder and Silverman model [6], is crucial for developing
in order to overcome these problems. After then, this personalized instructional strategies and content delivery
system creates customized recommendations based on mechanisms.
user choices in an effort to improve learning results. In
order to keep enhancing the system's efficacy, the study This model categorizes learning preferences based on
have also included a feedback mechanism and dimensions like active/reflective and visual/verbal,
performance evaluation. providing a robust framework for personalized learning
applications.
Keywords:- Personalized Learning, Mobile Learning,
Recommendation Systems, user Preferences, Learning In light of this, the purpose of this study is to
Experiences, Student Engagement, Adaptive Learning, investigate how to develop a customized mobile learning
Learning Styles, Educational Technology. system that takes into account the diverse character of
learners as well as the ever-changing educational
I. INTRODUCTION environment. This system creates adaptive, interesting, and
productive learning experiences across a range of topics and
Within the ever-changing field of education, mobile educational levels by utilizing data analytics, pedagogical
learning, or m-learning, has become a disruptive force that is theories, and emerging technologies.
redefining traditional learning environments and
pedagogical techniques. According to O'Malley et al. [1], m- II. REVIEW OF RELATED WORK
learning is any type of learning that takes place outside of a
set location and makes use of mobile technologies to In this section, reviewed literature review related to
enhance learning chances. This term emphasizes how Adaptive system, Personalized mobile learning system will
learners are fluidly mobile, interacting with a variety of be explored:
contexts and technologies to blur the boundaries between
formal and informal learning environments. In a study on the impact of adaptive learning
technology on student results, Johnson and Brown [9]
The proliferation of mobile devices such as emphasized the significance of customized approaches to
smartphones and tablets has empowered learners to access accommodate a range of learning preferences and styles.
educational resources anytime, anywhere, fostering a culture
of continuous and ubiquitous learning. This paradigm shift A study conducted by Thompson et al. [17] examined
in learning has profound implications, necessitating a how mobile learning interfaces affected students'
reevaluation of instructional design and delivery methods by engagement and retention of information, highlighting the
educators and institutions. importance of carefully designed interfaces in improving the
whole learning process.
Central to this paradigm shift is the acknowledgment
of learners' diverse preferences and characteristics. In their investigation into the use of recommendation
Traditional one-size-fits-all approaches are insufficient to algorithms to the customization of instructional materials,
address the varied needs of learners in the digital age. Garcia and Martinez [8] shown how these systems have the
Consequently, personalized learning has emerged as a capacity to adjust to individual preferences and enable
promising pedagogical approach that tailors’ educational customized learning opportunities.
experiences to individual learners' preferences, abilities, and
trajectories.

IJISRT24APR535 www.ijisrt.com 570

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

In a comparative examination of adaptive learning The integration of these components forms a robust and
systems, Lee et al. [12] evaluated the impact of these scalable system capable of delivering personalized learning
systems on student performance and emphasized the need experiences to users. See Figure 1.
for adaptive systems to continuously improve in order to
satisfy changing educational expectations.  User Profile Management:
User Profile Management serves as the foundation of
Wang and Chen [19] investigated the potential and the system, capturing crucial information about user
constraints of personalized mobile learning systems, preferences, learning styles, and contextual factors. A
addressing important issues including data security and user centralized database stores user profiles, which are updated
privacy while also spotting ways to improve personalized dynamically based on user interactions and feedback.
learning experiences through.
 Initial Questionnaire and Profile Updates:
 Related Works and Applications An initial questionnaire gathers user preferences during
With its adaptable learning platform, Khan Academy onboarding, allowing users to update their profiles over
has completely changed the face of online education. It was time. This component ensures that the system maintains
founded by Salman Khan and uses a mastery-based learning accurate and up-to-date user data for personalized
approach to provide individualized instruction in a range of adaptation.
areas. By adjusting difficulty levels and content dynamically
in response to learners' success, the platform allows for the  Context-aware Data Collection:
customization of the educational experience. Khan Context-aware Data Collection utilizes technologies
Academy's sophisticated adaptive algorithms are the reason like GPS and sensors to capture real-time user context,
behind its efficacy. The platform continuously assesses including location, environmental conditions, and device
students' performance as they use it, finding their strengths capabilities. This data informs adaptive content delivery and
and shortcomings so that upcoming sessions can be more user experience customization.
specifically tailored to them. Additionally, the system tracks
students' progress, providing them with information about  Learning Material Repository:
their strengths and areas for development. The methodology The Learning Material Repository houses a diverse
used by Khan Academy goes beyond the conventional one- range of educational content, categorized and tagged for
size-fits-all concept and makes education available to personalized recommendations. Content includes text-based
everyone. materials, multimedia resources, interactive exercises, and
assessments.
 Duolingo - Personalized Language Learning
The language-learning program Duolingo is an  Adaptation Engine:
example of how technology and individualized instruction The system's intelligent center is the Adaptation
may coexist. Its adaptive algorithms modify language Engine., employing recommendation algorithms, rule-based
lessons to the level, speed, and preferences of individual engines, and machine learning models to personalize
students. By using spaced repetition techniques, changing learning experiences. It integrates user profiles, contextual
focus based on user performance, and offering real-time data, and feedback to deliver tailored content and
feedback for an effective and entertaining learning process, interactions.
Duolingo gamifies the language learning experience. The
secret to Duolingo's success is its dedication to  User Interface:
individualized language instruction, which takes into The User Interface provides a user-friendly platform
account each learner's particular path. Duolingo is a prime for learners to access personalized content and engage with
example of individualized language learning, demonstrating learning materials. It incorporates intuitive design
the ability of technology to meet individual learning needs principles, interactive elements, and responsive layouts for
through the integration of adaptive technologies, seamless user experiences across devices.
gamification aspects, and spaced repetition tactics.
 Feedback and Iterative Improvement:
III. SYSTEM ARCHITECTURE Feedback mechanisms collect user input on suggested
materials and system performance, driving iterative
A comprehensive architectural design for a improvements to recommendation algorithms, content
personalized adaptive mobile learning system was selection, and user interface design. This continuous
developed. The system aims to revolutionize the educational feedback loop enhances the system's adaptability and
experience by dynamically tailoring learning content and effectiveness.
interactions based on individual user profiles and real-time
contextual data. Key architectural components include  Security and Data Privacy:
Context-aware Data Collection, User Interface, Feedback Robust security measures, including encryption
and Iterative Improvement, Adaptation Engine, User protocols, access controls, and compliance frameworks,
Interface Management, Security, Scalability and safeguard user data and ensure privacy. These measures are
Performance Optimization, and Continuous Monitoring and integrated across all components to maintain user trust and
Maintenance are some of the features that this system offers. regulatory compliance.

IJISRT24APR535 www.ijisrt.com 571

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

 Scalability and Performance Optimization:  System Processing: The system processes the user's
Scalability and Performance Optimization strategies, responses, applying the Felder-Silverman Learning
such as load balancing, caching, and cloud infrastructure, Model to categorize their learning style.
ensure that the system can handle increasing user  Tailored Recommendations: Leveraging the identified
interactions and data processing while maintaining learning style, the system generates personalized
responsiveness and reliability. recommendations for study materials, balancing visual,
verbal, active, and reflective, sequential elements.
 Conclusion:
The architectural design presented in this paper lays  Summary on Scenario 1:
the groundwork for a state-of-the-art personalized adaptive The learner, shaped by diverse preferences
mobile learning system. By integrating advanced encompassing visual, sensing, active, and sequential
technologies, user-centric design principles, and robust elements, engages in a tailored and effective learning
security measures, the system offers a personalized and experience, seeking clarity, practicality, collaboration, and
engaging learning experience, paving the way for future systematic understanding.
innovations in educational technology.
 Scenario 2: Interactive Learning Scenario
User Profile: A student preparing for exams.

 System Interaction:

 User Initiation: The user logs into the system and

voluntarily engages with the 12-question questionnaire
designed to uncover their learning preferences.
 Questionnaire Responses: The user provides responses to
questions addressing various dimensions, including
sensory preferences, information intake methods, and
problem-solving approaches.
 System Processing: The system processes the user's
responses, applying the Felder-Silverman Learning
Model to categorize their learning style.
 Tailored Recommendations: Leveraging the identified
learning style, the system generates personalized
recommendations for study materials, balancing visual,
verbal, active, and reflective, sequential elements.

 Summary on Scenario 2:
These recommendations aim to guide students
preparing for exams by aligning the learning strategies with
their individual preferences and styles.

IV. INTERVIEW AND DATA ANALYSIS

Fig 1 System Architecture and Design This part offers a synopsis of the interview-based study
methodology, describes the data analysis technique that was
 Scenarios used, and conducts a thorough examination of m-learning
preferences in relation to the learning styles of specific
 Scenario 1: Individual Learning Preferences learners. A single researcher guided 123 university students
Exploration who participated in the study on an individual basis. After
User Profile: A solitary user navigating the system for the interviews were taped, they were transcribed to allow for
personalized learning experiences. a thorough data analysis. To participate in the study, each
subject gave their informed consent. Each quiz took, on
 System Interaction: average, eight minutes to complete.

 User Initiation: The user logs into the system and One of the most important aspects of developing the
voluntarily engages with the 12-question questionnaire adaptive mobile learning system was training it with
designed to uncover their learning preferences. answers to a carefully constructed set of 12 questions. These
 Questionnaire Responses: The user provides responses to inquiries, which drew inspiration from the Felder-Silverman
questions addressing various dimensions, including learning styles model, served as the foundation for
sensory preferences, information intake methods, and understanding users' unique learning preferences, styles, and
problem-solving approaches. contextual inclinations.

IJISRT24APR535 www.ijisrt.com 572

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

Fig 2 Felder Silverman Learning Model

The 12 questions were carefully crafted to address a variety of aspects pertaining to learners' preferences; see Figure 3 for
more information.)

Fig 3 Diagram Showing the Questionnaire Obtained in Excel Format

IJISRT24APR535 www.ijisrt.com 573

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

The questions were divided into categories that about users' inclinations about characteristics like as Active
corresponded to the Felder-Silverman model's dimensions, versus Reflective, Sensing versus Intuitive, Visual versus
including visual/verbal, sensing/intuitive, sequential/global, Verbal, and Sequential versus Global. With the help of these
and active/reflective. Every question was thoughtfully data, users were categorized and the Personalized Adaptive
created to collect data that would help customize the user's Mobile Learning System was tailored to suit their preferred
learning experience. We have crafted our survey questions learning modes. When the system was being developed,
to delve into certain facets of the Felder and Silverman tokenization was essential to matching user inputs to Felder
Learning Styles Model, providing insightful information and Silverman Learning Styles Model sub.

Fig 4 Summary of the First three rows of user Questionnaire

The meticulous procedure went beyond simple word-level analysis and probed users' preferred methods of learning in order
to extract the nuances present in their answers. Figure 5 presents an overview of the distribution of learning styles based on the
survey processing phase.

Fig 5 Mapping user Expressions to the Sub Categories within the Felder and Silverman Learning Styles Model

IJISRT24APR535 www.ijisrt.com 574

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

The Felder and Silverman Learning Styles Model was and productive. The suggestions are based on an
used to carefully process the survey responses, tokenizing examination of user answers to an extensive survey that
them and classifying them into subgroups. This distribution looks at preferences related to different aspects of learning.
provides important information about the most common By utilizing the knowledge gathered from this data, the
learning styles within the user population. Greater frequency system creates unique interactive tasks for every user,
in particular categories indicates prevailing preferences, resulting in a flexible and captivating educational process. In
which form the basis of our Personalized Adaptive Mobile order to improve the user's educational experience, the
Learning System's later personalization and adaptation recommendations are tailored to individual preferences in
procedures. Comprehending this distribution is essential to processing, perception, input, and understanding. It has
customizing content recommendations that align with the individualized learning suggestions based on the Felder-
wide range of user learning preferences. See the Silverman paradigm for each user. Every participant has
visualization in Figure 6 below been allocated a User ID, which ranges from 1000 to 1123,
allowing for customized and individualized learning. The
utilization of customized advice guarantees an enhanced and
captivating learning experience for every user, cultivating an
adaptive mobile learning system that is unique to them.
Refer to figure 6.

V. PERFORMANCE EVALUATION

An extensive performance assessment and feedback

mechanism have been smoothly incorporated into the
system to gauge user happiness and the efficacy of the
personalized mobile learning system in place. This all-
encompassing methodology incorporates both quantitative
and qualitative evaluations, offering a thorough
understanding of the system's impact on its users.

 General Performance Satisfaction:

On a scale of 1 to 5, users indicated how satisfied they
were with the learning recommendation system and how
well it fulfilled their educational needs. Figure 7 shows this
quantitative evaluation and the effectiveness of the system's
performance.

Fig 6 A Visual Depiction of the 123 users'

Learning Preferences

Tokenization was employed by the system to enhance

its comprehension of user expressions, hence facilitating an
intricate alignment with the intricacies of the Felder and
Silverman Learning Styles Model. This method made a
substantial contribution to the system's ability to provide
users with tailored and adaptable learning recommendations
based on a deep understanding of their preferred learning
styles.

The research, which entailed a painstaking procedure

focusing on replies obtained from a set of carefully
constructed questions, was used to define learning style
classifications. The basis for classifying users according to
their preferred learning styles was provided by these Fig 7 Overall Performance Evaluation
carefully crafted questions that were divided into four
primary categories: processing, input, perception, and  Alignment of Recommendations with Learning
understanding. This method of classifying learning styles Objectives:
guarantees a detailed and personalized comprehension of the The main focus of the evaluation was how well
user's learning preferences. The system may personalize suggested learning resources matched users' learning
material recommendations that align with the most popular objectives. This evaluation measures how well the system
learning style by taking into account the majority response can adjust to users' evolving educational needs. For more
within each category. This makes learning more interesting information, see Figure 8.

IJISRT24APR535 www.ijisrt.com 575

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

 Gathering Additional Comments and Feedback:

In order to add depth to the review with a variety of
user views, users were urged to offer any extra comments or
feedback regarding their experience with the system.

Fig 8 Relevance of Recommendations of Learning Goals

 User-Friendliness and Usability:

Users evaluated the system's usability and convenience
of use, providing feedback on areas for improvement and
the efficacy of the user interface.

Fig 11 Collection of Additional Comments and Feedback

 A Customized Mobile Learning (m-Learning)

Application Tailored to Individual m-Learning
Preferences.
There are two stated instances where students have
different preferences for m-learning.

VI. CONCLUSION AND FUTURE WORKS

In order to determine the wide range of learning

preferences among the 123 participants surveyed, this study
examined customized learning using the Felder-Silverman
Learning Model. The Felder-Silverman categories analysis
of the responses gave important information about the most
Fig 9 Ease of use and User-Friendliness common learning styles. Customized recommendations are
made available to newcomers via an easy-to-use Google
 User-Generated Improvement Suggestions: Form, showcasing the project's potential to revolutionize
Users provided specific suggestions for improving the teaching methods. This approach promotes inclusion and
system, providing in-depth analysis of user expectations and active engagement by fusing user-friendly interfaces with
possible areas for development. data-driven insights. Future developments could incorporate
machine learning techniques for automated and improved
learning recommendations. Furthermore, expanding the
dataset and carrying out a more thorough user survey may
improve the learning model's resilience and applicability.
Working together with academic institutions or platforms,
we might implement and assess the system in real-world
settings, providing insightful information and confirming the
effectiveness of our recommendations for individualized
learning. Frequent system upgrades and improvements
based on user feedback and changing pedagogical
approaches would guarantee the system's applicability and
flexibility in a changing educational environment.

Fig 10 User-Defined Suggestions for Improvement

IJISRT24APR535 www.ijisrt.com 576

Volume 9, Issue 4, April – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24APR535

REFERENCES [18]. Turner, L., et al. (2019). Long-term Effects of

Personalized Learning Interventions on Academic
[1]. Brown, R., & Jackson, M. (2019). Advantages and Achievement. Educational Psychology Review,
Drawbacks of Mobile Learning Platforms. 27(3), 401-415.
International Journal of Educational Technology, [19]. Wang, J., & Chen, L. (2021). Challenges and
18(1), 56-68. Opportunities in Personalized Mobile Learning
[2]. Chen, Z., et al. (2020). Global Context-Aware Systems. International Journal of Educational
Progressive Aggregation Network for Salient Object Technology, 14(2), 45-58.
Detection. AAAI Conference on Artificial [20]. Zhang, Q., et al. (2020). Meta-analysis of Adaptive
Intelligence, 1(2), 123-135. Learning Systems on Student Performance. Journal
[3]. Dai, Z., & Callan, J. (2019). Context-Aware of Educational Technology, 50(2), 189-202.
Sentence/Passage Term Importance Estimation for
First Stage Retrieval. arXiv.Org, 1(1), 34-47.
[4]. Dey, A., et al. (2001). Conceptual Framework and
Toolkit for Context-Aware Applications. Human-
Computer Interaction, 16(3), 166-197.
[5]. Duolingo (2017). Personalized Language Learning
with Adaptive Algorithms. Journal of Language
Education, 25(1), 34-47.
[6]. Elayan, H., et al. (2021). Digital Twin for Intelligent
Context-Aware IoT Healthcare Systems. IEEE
Internet of Things Journal, 8(5), 16749-16757.
[7]. Felder, R., & Silverman, L. (1988). Learning and
Teaching Styles in Engineering Education.
Engineering Education, 78(7), 674-681.
[8]. Garcia, D., & Martinez, M. (2020). Role of
Recommendation Algorithms in Tailoring
Educational Content. International Journal of E-
Learning, 12(4), 123-135.
[9]. Johnson, A., & Brown, B. (2018). Impact of
Adaptive Learning Technologies on Student
Outcomes. Educational Technology Research and
Development, 66(3), 689-702.
[10]. Khan, S. (2015). Khan Academy: Revolutionizing
Online Education with Adaptive Learning. Journal of
Online Learning, 10(2), 78-92.
[11]. Kim, Y., et al. (2018). Aligning Adaptive Learning
Approaches with Cognitive Science Principles.
Journal of Cognitive Education, 35(3), 321-335.
[12]. Lee, H., et al. (2019). Comparative Analysis of
Adaptive Learning Systems on Student Performance.
Journal of Educational Technology, 48(1), 89-102.
[13]. O’Malley, C., et al. (2003). Definition and
Characteristics of Mobile Learning. Journal of
Mobile Learning, 5(2), 45-56.
[14]. Patel, A., & Gupta, S. (2022). Integration of
Artificial Intelligence in E-Learning
Recommendation Systems. Journal of Artificial
Intelligence in Education, 28(4), 567-580.
[15]. Schilit, B., et al. (1994). Context-Aware Computing
Applications. First Workshop on Mobile Computing
Systems and Applications, 1(1), 85-90.
[16]. Smith, J., & Johnson, K. (2018). Gamification and
Spaced Repetition in Language Learning: The Case
of Duolingo. Journal of Educational Technology,
42(3), 213-226.
[17]. Thompson, C., et al. (2017). Enhancing Student
Engagement and Knowledge Retention through
Mobile Learning Interfaces. Journal of Educational
Technology, 45(2), 256-269.

IJISRT24APR535 www.ijisrt.com 577