Volume 9, Issue 12, December – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

The Role of Adaptive AI in Skill

Development: A Comprehensive Survey
Sowmyashree S1; Manas Shet N2; Prashanth D M3; Pramanth P A4; Raghunandana M S5
Department of Information Science RNS Institute of Technology, Bengaluru

Abstract:- E-learning systems facilitate accessible, II. LITERATURE SURVEY

flexible learning but often lack adaptive features tailored
to individual student needs, leading to reduced The literature on adaptive and intelligent tutoring
effectiveness. Adaptive Intelligent Tutoring Systems systems (AITS) in e-learning explores various approaches to
(AITS) and Adaptive Hypermedia (AH) offer personalize learning, improve engagement, and enhance
personalized support by monitoring student progress educational outcomes. The studies on AITS, adaptive
and adapting content and instruction. However, AITS hypermedia, cognitive modeling, and AI-powered tutoring
excels in customized guidance, while AH generally are:
provides technical resources without the same level of
adaptation. The AI-Tutor system proposes integrating  Integration of AITS and Adaptive Hypermedia (AH):
AITS and AH into an Adaptive Intelligent Tutoring The integration of Intelligent Tutoring Systems (AITS)
System (AITS) to create a more effective e-learning and Adaptive Hypermedia (AH) to create Adaptive
environment. This combined approach promotes Intelligent Tutoring Systems (AITS) aimed at enhancing e-
personalized knowledge sharing and enhances students' learning. It critiques traditional e-learning approaches that
learning outcomes through adaptive, interactive content. lack adaptability, suggesting that AITS can personalize
content based on learners' needs, improving engagement and
Keywords:- E-learning Platform, Adaptive Intelligent learning outcomes by dynamically adjusting to each
Tutoring Systems (AITS), Adaptive Hypermedia (AH), student's knowledge and learning style.
Personalized Instruction, Adaptive Learning, and Knowledge
Sharing.  User-Centered Adaptation in Human-Computer
Interaction (HCI):
I. INTRODUCTION Benyon and Murray focus on adaptive systems in
Human-Computer Interaction (HCI) that respond to user-
E-learning enables flexible learning, fostering specific cognitive and personality through AITS. Through
motivation, communication, and skill acquisition through user modeling, they propose that adaptive systems tailored
accessible, internet-based technology. By removing physical to individual cognitive abilities can improve interface
distance, e- learning connects students and instructors usability. Their findings underscore the need for user-
virtually, facilitating knowledge transfer and personal centered adaptations, suggesting that future educational
development across education, business, and training systems should account for cognitive diversity in interface
contexts. AITS has benefits, which include cost- design to support effective learning.
effectiveness, time savings, and measurable learning
outcomes, making it a practical approach to education.  Paula J. Durlach, Alan M. Lesgold-2012:
The Tutor System discusses the state of adaptive
However, traditional e-learning often lacks the training technology, particularly in structured domains like
adaptability needed to cater to individual learning needs, algebra. It highlights the challenges of adapting these
which can reduce AITS effectiveness. Intelligent Tutoring technologies for less-structured scenarios, such as
Systems (AITS) and Adaptive Hypermedia (AH) have simulations and serious games. Contributions within the
emerged as solutions, offering adaptive instruction and volume cover topics such as student modeling and virtual
tailored content through real- time student performance agents, making it relevant for developers and researchers
monitoring. AITS provides personalized guidance, while AH aiming to expand adaptive systems to diverse educational
delivers technical resources, making both technologies and professional settings.
integral to improving e-learning experiences.
 Wenbin Gan, Yuan Sun – 2019:
To address these limitations, The integration of Gan and Sun introduce "AI-Tutor," an AI-powered
Adaptive Intelligent Tutoring System (AITS) for e-learning. adaptive tutoring system that provides personalized support
AITS enables content adaptation based on each student’s by analyzing cognitive assessments to target each student’s
needs, facilitating a more interactive and effective learning weak points. The system generates tailored questions and
environment. This integration promotes dynamic, feedback, aiming to improve knowledge retention
personalized learning, allowing students to interact with and efficiently.
apply new knowledge more meaningfully. supports
constructivist learning principles and enhances educational
outcomes.

IJISRT24DEC234 www.ijisrt.com 134

Volume 9, Issue 12, December – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
 Dr. Adeel Baig, Dr. John D. Cressler, Dr. Marvin  Enhance Learning Accessibility and Flexibility: Offer
Minsky-2024: a flexible learning platform that students can access at
The system discusses the role of AI in Indonesia’s their convenience, eliminating time and location
digital transformation, focusing on AITS application across constraints associated with traditional education.
sectors such as education, healthcare, and agriculture. It  Measure and Improve Learning Outcomes:
identifies both opportunities and challenges, including Implement analytics and tracking tools to assess the
regulatory, ethical, and infrastructure issues, emphasizing effectiveness of personalized learning strategies,
the importance of robust AI frameworks. Although not continuously refining the system based on student
solely education-focused, the insights on AI's transformative performance and feedback.
impact provide context for implementing AI-driven adaptive  Support Collaborative Learning Experiences: Provide
learning in broader educational reforms. platforms for students to collaborate with peers and
instructors, fostering a community-driven approach to
 Ambroise Baillifard, Maxime Gabella–2023: knowledge sharing and problem-solving.
Baillifard and Gabella examine AI tutors in educational
settings, emphasizing personalization and adaptive learning  Existing System
techniques like spaced repetition. In a study at UniDistance Traditional e-learning platforms are web-based
Suisse, an AI tutor modeled students' progress and systems that provide access to course materials and resources
implemented retrieval- based practices, resulting in higher but often rely on static content like lecture notes and videos.
academic performance. Their research validates the role of These systems typically lack interactivity and fail to adapt to
neural-network-driven AI tutors in improving learning individual learner preferences or progress. Adaptive
outcomes by adapting to individual learning paces and needs, Hypermedia (AH) offers some customization but focuses
reinforcing the effectiveness of personalized e- learning mainly on teaching concepts rather than problem-solving.
strategies. Intelligent Tutoring Systems (AITS) excel in problem-
solving but struggle to cover diverse knowledge areas. The
 Rainer Winkler, Julian Roos - 2019: lack of personalization and adaptability in current e-learning
The research identifies key design principles for SPAs, solutions highlights the need for more advanced, learner-
including interactivity, scaffolding, humanness, centered approaches.
engagement, transparency, accessibility, flexibility, response
time, and supportiveness. The system focuses on group  Proposed System
discussions and a quasi-field experiment to evaluate the The AI-Tutor System offers a deeply customized
effectiveness of their designed SPA, which shows significant learning experience by integrating cognitive assessments
improvement in learning outcomes. and custom question generation to suit each student’s needs.
By combining adaptive course content with an automated
 Objectives problem-solving feature, the system provides targeted
The goal of this survey is to thoroughly examine and support that aligns with individual learning gaps. Through
compare current machine learning methods used for AI advanced machine learning and natural language processing,
Tutor Systems. AI-Tutor evaluates students’ knowledge levels and creates
personalized questions that address specific areas for
 Tailor Educational Content to Individual Learning improvement, boosting retention and engagement. The
Needs: Develop a system that customizes learning system framework includes adaptive materials, tailored
materials and instructional approaches based on each learning paths, and a data-driven problem solver that guides
student's unique learning preferences, performance, and students with step-by-step feedback. This setup fosters a
progress. dynamic, interactive learning environment that adapts in real
 Enhance Engagement and Motivation: Integrate time to each learner’s progress, supporting continuous
interactive and adaptive features that encourage student improvement through ongoing analytics.
participation and keep them motivated throughout their
online learning experience.  Advantages of Proposed System
 Monitor and Adjust Student Progress in Real- Time:
Enable both instructors and the system to track students'  Customized Learning: The system creates
progress in real-time, adapting the content and delivery individualized learning paths and resources, using
to suit their evolving learning needs. cognitive assessments to pinpoint and address each
 Incorporate Constructivist Learning Strategies: student’s specific needs, improving overall learning
Provide simulations and real-world challenges that effectiveness.
promote active learning, allowing students to apply  Instant Feedback and Flexibility: The system actively
knowledge through hands-on problem- solving. tracks student progress, allowing it to give immediate
 Facilitate Long-Term Knowledge Retention: Design feedback and adjust content dynamically, enhancing the
learning paths that adapt to students’ needs, offering learning experience in real time.
personalized feedback and ensuring that they can retain  Focused Remediation: AI-Tutor offers personalized
and effectively apply what they have learned. question sets and detailed problem- solving guidance,
helping students strengthen their understanding of topics
efficiently compared to traditional approaches.

IJISRT24DEC234 www.ijisrt.com 135

Volume 9, Issue 12, December – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
 Enhanced Problem-Solving Skills: With an AI-  Combination of AITS and Adaptive Hypermedia (AH):
powered solver, students receive step-by-step hints for Researchers began merging AITS with Adaptive
challenging problems, promoting independent thinking Hypermedia (AH), as AH organizes online learning
and a deeper grasp of concepts. materials into accessible hypertext. This combination led to
 Adaptive Learning Paths: Using data analytics, the the development of Adaptive Intelligent Tutoring Systems
AI-Tutor creates tailored course sequences and materials, (AITS), which leverages AITS and AH to deliver adaptive,
making learning more engaging and aligned with each personalized learning.
student’s pace.
 Expert Model with Adaptive Presentation and
III. METHODOLOGY Navigation:
The expert model within AITS includes Adaptive
 The Methodology of the Proposed System Involves Presentation and Adaptive Navigation Support, allowing it
Several Stages: to be applicable across various domains. This design enables
the modification of lecture content, decision rules, and the
 Collecting Learner Data: The system first gathers evaluation criteria within the expert model, allowing AITS
essential details about each learner, such as their to adapt to diverse learning contexts.
preferred learning style, current knowledge level,
objectives, and interests. This information is used to  Content Personalization Based on Learner Knowledge:
develop a tailored learning path and choose the most The expert model uses stored information about the
suitable resources for them. learner’s knowledge to recommend relevant content. It
 Creating a Personalized Learning Path: Using assesses what a learner already knows and what they need to
diagnostic models, the system evaluates the learner’s learn next, using prerequisite relationships between topics to
understanding of the subject and identifies areas that suggest suitable lectures. The learner can either be shown a
need improvement. Based on this evaluation, it crafts a recommended lecture or given a choice based on their
customized learning path with specific content to meet learning style.
the learner’s needs.
 Tracking Progress in Real Time: As the learner moves  Dynamic Selection of Learning Material:
through the material, the system logs data on completed By blending AH and AITS, AITS can adaptively select
courses, types of exercises attempted, and test results. the most suitable learning material from AITS knowledge
This continuous tracking helps the system adapt content base, presenting it at the ideal time and in the right format
tothe learner's progress and performance. for each learner. This method ensures that all educational
 Generating Customized Questions: After the learner resources are used efficiently, providing targeted support for
completes a topic, the system creates specific questions concept mastery and practical application.
to address areas where they may need more practice,
ensuring targeted reinforcement of weaker points.  Integrated Instructional Approaches for Comprehensive
Learning:
 Providing Step-by-Step Problem Solving and
Support: The system includes an AI-driven problem Since AH is ideal for conveying conceptual knowledge
and AITS excels in problem-solving support, combining
solver that offers hints and detailed solutions as learners
work on questions. This support promotes independent both approaches offers a comprehensive learning
experience. Together, they allow AITS to adapt content
learning by guiding them through complex problems
dynamically according to each learner’s behavior,
with structured assistance.
preferences, and progress online.
IV. SYSTEM ARCHITECTURE

 Content and Knowledge Outside the System:

Originally, Intelligent Tutoring Systems (AITS) relied
on external learning resources, like lectures or textbooks, for
foundational knowledge. However, as digital capabilities
grew, combining AITS with online learning materials in one
integrated system became more effective.

IJISRT24DEC234 www.ijisrt.com 136

Volume 9, Issue 12, December – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 1 Conceptual of Adaptive Intelligent Tutoring Systems for E-Learning System

V. CONCLUSION [5]. R. Soltanpoor, C. Thevathayan, and D.

D’Souza, “Adaptive remediation for novice
Adaptive Intelligent Tutoring Systems (AITS) is the programmers through personalized prescriptive
valuable tools for computer-based learning. Current efforts quizzes”, 2018.
focus on merging these systems to enhance e-learning by
[6]. David Benyon, Dianne Murray, “Adaptive
providing personalized instruction that adapts to each
student’s learning style. AITS acts as an integrated model, Systems: from intelligent tutoring to
allowing both systems to share data on a learner’s progress. autonomous agents ,” 1993.
This approach aims to strengthen the connection between [7]. Paula J. Durlach, Alan M. Lesgold, “Adaptive
teaching concepts and skill application, ultimately Technologies for Training and Education”,
improving learning outcomes. 2012.
[8]. Wenbin Gan, Yuan Sun, “AI-Tutor: Generating
REFERENCES Tailored Remedial Questions and Answers
Based on Cognitive Diagnostic Assessment”,
[1]. Pipatsarun Phobuna, Jiracha Vicheanpanya, 2019.
“Adaptive intelligent tutoring systems for e- [9]. Y. Sun, M. Suzuki, and S. Kakinuma, “Effective
learning systems”, 2010. feedback for selfregulated learning: Applying
[2]. P. Liao, Y. Sun, S. Ye, X. Li, G. Su, and Y. cognitive diagnostic assessment”, 2012.
Sun, “Predicting learners’ multi-question [10]. Rainer Winkler, Julian Roos, “Bringing AI into
performance based on neural networks” , 2017. the Classroom: Designing Smart Personal
[3]. D. S. Chaplot, E. Rhim, and J. Kim, Assistants as Learning Tutors”, 2019.
“Personalized adaptive learning using neural [11]. Dr. Adeel Baig, Dr. John D. Cressler, Dr.
network”, 2016. Marvin Minsky, “The Future of AI in
[4]. Kenneth R. Koedinger, Emma Brunskill, Ryan Education: Personalized Learning and
S. J. d. Baker, Elizabeth A. McLaughlin, “New Intelligent Tutoring Systems”, 2024.
Potentials for Data-Driven Intelligent Tutoring [12]. Y. Sun, S. Ye, S. Inoue, and Y. Sun,
System Development and Optimization”, 2013. “Alternating recursive method for q-matrix
learning”, 2014.

IJISRT24DEC234 www.ijisrt.com 137

Volume 9, Issue 12, December – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
[13]. C. Wongwatkit, N. Srisawasdi, G.-J. Hwang,
and P. Panjaburee, “Influence of an integrated
learning diagnosis and formative assessment-
based personalized web learning approach on
students learning performances and perceptions”,
2017.
[14]. Yu, M. Wang, W. Gan, B. He, and N. Ye, “A
framework for solving explicit arithmetic word
problems and proving plane geometry theorems”,
2018.
[15]. Ambroise Baillifard, Maxime Gabella,
“Implementing Learning Principles with a
Personal AI Tutor: A Case Study”, 2023.

IJISRT24DEC234 www.ijisrt.com 138