Project Title: AI-Powered Personalized Python Learning Platform:
An Intelligent Tutoring System for Adaptive, Real-Time
Programming Education
Team no: 18
Team Members:
AYUSHMAN SHARMA 1NT22IS032
GOVIND KUMAR 1NT22IS055
AMAN RAJ 1NT22IS010
1)Introduction:-
The integration of Artificial Intelligence (AI) in education has revolutionized the way learners
engage with academic content. With the growing importance of programming skills, particularly
Python, there is a pressing need for adaptive and intelligent systems that can provide
personalized, real-time support to students. Traditional online learning platforms often lack
dynamic feedback, customization, and interactive guidance, which can result in disengagement
and poor learning outcomes.
Our project, "AI-Powered Personalized Python Learning Platform: An Intelligent Tutoring
System for Adaptive, Real-Time Programming Education," aims to bridge this gap by
building a smart educational assistant tailored for Python learners. Drawing inspiration from
existing AI-based platforms and tutoring systems, our solution incorporates chatbot-based
assistance, automated assessments, and adaptive learning paths. By leveraging AI technologies
such as Natural Language Processing (NLP), machine learning, and large language models, the
platform will provide real-time responses to queries, identify learning gaps, and recommend
targeted learning content.
The motivation for choosing this topic stems from the promising results observed in several
research studies that demonstrate improved learning outcomes, engagement, and personalization
through AI-enhanced tools. Our platform aims to combine these capabilities into a focused
Python learning system to support beginners and intermediate learners in mastering
programming concepts effectively.
2)Literature Survey:-
2.1 Paper 1 Summary:-
Paper Title / Source:
Title: AI-Enhanced Learning Assistant Platform
Published In / Year: IEEE, Proceedings of the 7th International Conference on Inventive
Computation Technologies (ICICT), 2024
Link to the Paper: IEEE Xplore DOI
Key Contributions / Findings:
This paper introduces a comprehensive AI-powered educational platform that integrates key
features like automated question generation, answer evaluation, weak area detection, recursive
testing, a query forum, and expert chat support. It leverages NLP and machine learning to
generate content-specific questions, provide instant feedback, and track individual student
weaknesses. The recursive testing approach with spaced repetition reinforces long-term
retention. An integrated peer-to-peer and expert-supported forum further encourages
collaborative learning.
Limitations Identified:
● Q&A generation relies heavily on the provided content, risking limited topic coverage.
● Feedback from the system may be inaccurate or misleading due to evaluation flaws.
● Weak area detection can be imprecise, limiting the effectiveness of tailored support.
● Recursive testing can cause boredom and reduced motivation.
● Response time and availability of expert support is not guaranteed.
● Lack of adaptability to diverse learning styles like visual or hands-on methods.
Relevance to Your Project:
The platform’s focus on adaptive feedback, personalized content, and collaborative tools directly
aligns with the goals of developing an AI-powered learning assistant. It supports the concept of
AI replacing or enhancing human tutors.
Research Gap Identified:
There is a need for:
● More advanced NLP for better content coverage.
● Robust feedback mechanisms.
● Improved personalization and multimodal learning support.
● Enhanced support responsiveness and scalability.
2.2 Paper 2 Summary:-
Paper Title / Source:
Title: AI-Enhanced Learning Assistant Platform: An Advanced System for Q&A Generation
from Provided Content, Answer Evaluation, Identification of Students' Weak Areas, Recursive
Testing for Strengthening Knowledge, Integrated Query Forum, and Expert Chat Support
Published In / Year: IEEE, 2nd International Conference on Artificial Intelligence and Machine
Learning Applications (AIMLA), 2024
Link to the Paper: IEEE Xplore DOI
Key Contributions / Findings:
This study presents an upgraded AI-Enhanced Learning Assistant Platform designed to
overcome limitations of traditional learning systems by implementing features such as
content-based Q&A generation, answer evaluation, weak area detection, recursive testing, a
student query forum, and real-time expert chat. The platform uses NLP and ML to create
personalized learning experiences and supports adaptive pathways for different learning needs.
Feedback mechanisms and spaced repetition reinforce knowledge, while integrated chat and
forums promote collaboration.
Limitations Identified:
● Q&A generation still suffers from dependence on the initial content quality.
● Evaluation feedback may not always be accurate or insightful.
● Weak area detection can be inconsistent.
● Recursive testing may feel repetitive and boring for students.
● The chat support may lag or lack scalability due to limited human expert availability.
● Insufficient accommodation for diverse learning styles (e.g., visual, hands-on learners).
Relevance to Your Project:
This platform aligns directly with the AI-powered assistant concept, incorporating intelligent
features aimed at personalizing and enhancing the learning process, particularly through adaptive
feedback and student-specific interventions.
Research Gap Identified:
● Need for deeper semantic understanding for Q&A generation.
● More advanced and contextual answer evaluation.
● Improved adaptability for diverse learners.
● Scalable expert systems (possibly AI-driven) for consistent support.
2.3 Paper 3 Summary:-
Paper Title / Source:
Title: Student Preferences in Interacting with AI-Enhanced Learning Assistants (AIELA): A
Comparative Study
Published In / Year: IEEE, SoutheastCon 2025
Link to the Paper: IEEE Xplore DOI
Key Contributions / Findings:
This study evaluates student interaction with two versions of the AI-Enhanced Learning
Assistant (AIELA)—a Raspberry Pi hardware-based model and a modern web-based React
Native version. AIELA uses GPT-4-turbo for generating responses to learning queries. The study
explores student satisfaction, usability, and engagement through classroom demos and surveys.
Results show that students found the web-based version easier to use and more accessible,
scoring it moderately on effectiveness (3.81/5) and comfort (4.15/5). However, deeper
conceptual learning and misconception resolution scored lower (~2.96/5). Compared to human
learning assistants (HLAs), AIELA was preferred for convenience but fell short in pedagogical
depth.
Limitations Identified:
● Lacks depth in clarifying misconceptions and supporting deeper learning.
● Some students were frustrated by the assistant’s Socratic approach (indirect feedback).
● AI’s refusal to give direct answers led to students attempting to “jailbreak” the system.
● Limited conceptual understanding support when compared to human tutors.
● Effectiveness is affected by students’ expectations and previous experience with more
powerful tools like ChatGPT.
Relevance to Your Project:
AIELA serves as a real-world AI-powered learning assistant model, offering insights into
technical design (web vs. hardware interface), student engagement, and the role of AI in active
learning environments—closely aligning with your project’s goals.
Research Gap Identified:
● Need for improving the assistant’s depth in handling complex topics.
● Better balance between Socratic guidance and direct instruction.
● Advanced conversational strategies and longer-term classroom trials recommended for
future development.
2.4 Paper 4 Summary:-
Paper Title / Source: Implementation of AI-driven technology into education – a pilot study
on the use of chatbots in foreign language learning
Published In / Year: Cogent Education, 2024
Link: https://doi.org/10.1080/2331186X.2024.2355385
Key Contributions / Findings: Demonstrated improved foreign language skills (English, B2-
C1) in students using AI chatbots. Students showed a positive attitude towards chatbots and
reported increased confidence and engagement.
Limitations Identified: Study limited to one-month duration and one subject (foreign language)
with a small sample size.
Relevance to Your Project: Supports the utility of chatbots in real-world classroom settings,
particularly for language acquisition.
Research Gap Identified: Lacks evaluation across subjects and long-term academic impact.
2.5 Paper 5 Summary:-
Paper Title / Source: Exploring the prospects of using artificial intelligence in education
Published In / Year: Cogent Education, 2024
Link: https://doi.org/10.1080/2331186X.2024.2353464
Key Contributions / Findings: Personalized AI learning significantly enhanced knowledge in
English, computer science, economics, and anatomy. Students and teachers rated the AI
learning process positively.
Limitations Identified: AI dominance led to slightly reduced student engagement. No
comparative human-based instructional strategies studied.
Relevance to Your Project: Validates AI’s adaptability to multiple disciplines and personalized
learning potential.
Research Gap Identified: Further exploration needed on emotional engagement and ethical
design of AI tools.
2.6 Paper 6 Summary:-
Paper Title / Source:
Chatbots in education: Hype or help? A meta-analysis
Published In / Year: Learning and Individual Differences, 202
Link: https://doi.org/10.1016/j.lindif.2025.102646
Key Contributions / Findings: Meta-analysis of 62 studies found chatbots have a statistically
significant small to moderate effect on learning performance. Most effective in STEM fields,
lower education levels, and text-based interactions.
Limitations Identified: Potential publication bias. Limited studies on voice-based bots and
advanced LLMs.
Relevance to Your Project: Provides a broad evidence base for educational chatbot benefits.
Research Gap Identified: Cultural variations, long-term effects, and platform-specific
analyses remain underexplored.
2.7 Paper 7 Summary:-
Paper Title / Source: Understanding AI Chatbot adoption in education: PLS-SEM analysis
of user behavior factors
Published In / Year: Computers in Human Behavior: Artificial Humans, 2024
Link: https://doi.org/10.1016/j.chbah.2024.100098
Key Contributions / Findings: Used PLS-SEM modeling to link optimism, interaction, and
accuracy to perceived ease of use and intention to adopt chatbots. Ethics and insecurity were
found to hinder adoption.
Limitations Identified: Did not evaluate actual academic performance, only intention and
perception.
Relevance to Your Project: Informs the behavioral side of chatbot use in education, critical for
designing user-accepted tools.
Research Gap Identified: Future research must assess real performance outcomes and expand
to teacher adoption models.
2.8 Paper 8 Summary:-
Paper Title/Source: Development and Evaluation of an AI-Enhanced Python Programming
Education System
Published In / Year: IEEE UEMCON / 2024
Link: IEEE Xplore Link
Key Contributions: This paper outlines the design of an AI-powered Python learning platform
built around GPT-3.5 Turbo. It features three core components: a question-answering chatbot, a
diagnostic test for personalized course recommendations, and an AI-based quiz grading system.
The study reported high performance, with 28/30 consistency in grading and 26/30 accuracy in
chapter recommendations. The chatbot was particularly effective in boosting learner engagement
by providing instant support.
Limitations: Feedback delivery was inconsistent, content coverage was incomplete (only two of
six chapters implemented), and the system experienced latency with API calls.
Relevance to This Project: The paper offers a valuable framework for chatbot integration and
feedback mechanisms.
Identified Gaps: There is a need for full content coverage, improved feedback reliability, and
faster AI response times.
2.9 Paper 9 Summary:-
Paper Title/Source: Exploring the Role of AI-Based Tools in Enhancing Student Learning
Outcomes
Published In / Year: IEEE ICEED / 2024
Link: IEEE Xplore Link
Key Contributions: This empirical study investigated how AI tools affect learning among
first-year university students. It highlighted significant improvements in student engagement,
autonomy, and academic performance. Notably, 94% of students reported that AI tools made
assignments easier, while 86% noticed improvement in their academic results.
Limitations: The study was largely based on self-reported data, lacking long-term metrics or
control over external factors.
Relevance to This Project: The findings confirm student preference for AI-driven, personalized
learning solutions and support the integration of adaptive feedback systems.
Identified Gaps: There was limited objective evaluation and insufficient comparison of AI
versus human-driven feedback quality.
2.10 Paper 10 Summary:-
Paper Title/Source: Introducing a Smart Learning Assistant on a MOOC Platform
Published In / Year: IEEE GECon / 2023
Link: IEEE Xplore Link
Key Contributions: This study introduced a Smart Learning Assistant designed for a MOOC
platform using RASA. The assistant was capable of answering course-related queries, guiding
learners through content, and suggesting study materials based on learner profiles. The
implementation of a feedback loop improved the assistant’s accuracy over time.
Limitations: Despite early promise, the assistant offered limited functionality and struggled with
complex queries and natural language understanding. It also lacked the domain-specific
intelligence needed to support a wide range of learners.
Relevance to This Project: The study provides design insights for developing intelligent
tutoring agents and highlights the importance of robust personalization.
Identified Gaps: Deficiencies in emotional intelligence, limited range of responses, and weak
support for diverse user inputs constrained the assistant’s effectiveness.
2.11 Consolidated Summary of All 10 Papers:-
The literature surveyed across ten peer-reviewed research papers from reputed publishers (IEEE,
Elsevier, Taylor & Francis) consistently highlights the transformative role of AI in enhancing
education, especially through intelligent tutoring systems, chatbots, and adaptive learning
platforms. These studies collectively demonstrate how AI can offer real-time feedback,
personalized learning paths, and improved engagement for students in diverse subjects including
language learning, STEM, and programming.
Several papers (e.g., Papers 1, 2, 8, 10) focused on comprehensive AI-enhanced learning
platforms that integrate features such as question generation, answer evaluation, weak area
detection, and recursive testing. These systems employed Natural Language Processing (NLP)
and Machine Learning (ML) techniques to deliver personalized and adaptive learning support.
Notably, Paper 8 directly addressed Python education, showing the effectiveness of GPT-based
systems in boosting learner performance and engagement.
Other works (e.g., Papers 3, 4, 5, 6) explored chatbot-driven learning environments. These
studies found statistically significant improvements in learner performance, confidence, and
satisfaction. Meta-analyses and case studies confirmed the value of AI assistants across multiple
educational settings, though limitations included short-term study durations, small sample sizes,
and underdeveloped emotional or conceptual support.
A behavioral perspective was provided in Paper 7, which analyzed user adoption of chatbots. It
highlighted factors like trust, accuracy, and perceived usefulness as critical to successful AI tool
implementation. Additionally, Papers 9 and 10 stressed the importance of user experience and
performance feedback, noting that real-time support, when designed well, can improve both
autonomy and academic results.
Despite their contributions, most studies revealed common gaps: limited subject-specific
implementations (especially in programming), lack of emotional intelligence in AI systems,
inconsistent feedback quality, and scalability challenges. There is also a notable absence of
long-term studies comparing AI tools with traditional teaching methods across diverse learner
groups.
In conclusion, the reviewed literature strongly supports the feasibility and impact of AI-powered
tutoring systems. These insights directly inform our proposed platform, which aims to build upon
these advancements by creating a personalized, real-time, Python-focused intelligent tutoring
system that addresses current limitations and enhances the learning experience.
3)Problem Statement:-
Despite the rise of digital education, most platforms for learning programming remain static and
generic. They often fail to adapt to individual learner needs, lack real-time personalized
feedback, and provide limited support for concept reinforcement. This is especially critical in
programming education, where timely clarification, practice, and conceptual depth are essential.
Research has shown that AI-powered systems can significantly enhance learning by offering
adaptive content, tracking learner performance, and simulating human-like tutoring. However,
current implementations suffer from limitations such as:
● Incomplete content coverage and weak handling of advanced concepts.
● Limited emotional intelligence or personalized engagement strategies.
● Inconsistent or delayed feedback mechanisms.
● Low adaptability to diverse learning styles or levels of expertise.
Our project seeks to solve these issues by developing a Python-focused intelligent tutoring
system that is scalable, interactive, and capable of delivering adaptive learning experiences in
real time. The system will act as a virtual assistant and tutor, helping learners progress efficiently
and confidently.
4)Objectives of the Project:-
The key objectives of this project are:
● Objective 1: To design and implement an AI-powered learning platform specifically for
Python programming that provides real-time query resolution, code assistance, and
concept explanations.
● Objective 2: To integrate personalized learning features such as weak area detection,
automated assessments, and adaptive content recommendation using machine learning
and natural language processing techniques.
● Objective 3: To evaluate the platform’s effectiveness in improving learning outcomes,
engagement, and satisfaction compared to traditional static content by collecting user
feedback and performance data.
5)Methodology for the Proposed System:-
The proposed project is a lightweight, AI-inspired Python learning assistant designed for offline
use. It combines quiz-based assessment, pseudo-AI chatbot logic, mood-based feedback, and
weak area tracking to deliver personalized support without the need for a backend or internet
access. This approach offers an innovative solution for learners with limited connectivity while
still benefiting from AI principles like adaptivity and personalization.
5.1 System Overview:-
The system consists of three primary modules:
1. Python Quiz Engine
2. Mood-Based Feedback & Suggestion System
3. Static Chatbot with Python Concepts (GPT-style logic using JSON)
Each module operates on locally stored JSON files and uses simple decision logic to simulate
intelligent responses and adaptability.
5.2 Modules in Detail:-
● Quiz Engine:
Users take small topic-based quizzes (loops, lists, functions, etc.). Each quiz result is
stored locally and used to identify weak areas.
● Mood Selector + Feedback System:
Before starting, users select their mood (e.g., confused, curious, frustrated). Based on
this, the app tailors the tone of the chatbot and feedback messages (e.g., motivational,
step-by-step).
● Chatbot Assistant (Static but Smart):
A local JSON-based chatbot simulates intelligent answers to Python FAQs (syntax,
examples, debugging). No OpenAI needed.
● Recommendations:
Based on quiz history and mood, the app suggests which topic to revisit and provides
practice problems.
5.3 Technology Stack:-
● Frontend: Flutter (for mobile) or HTML/JS (for web)
● Data: Local JSON files (e.g., quiz_data.json, chat_responses.json)
● Optional AI: Use GPT responses saved offline for chatbot answers
● No backend, no API calls – works 100% offline
5.4 Workflow:-
1. User launches app → selects mood.
2. Takes topic quiz → scores saved locally.
3. Weak areas tracked using basic rules (e.g., score < 50%).
4. Chatbot and feedback adjusted based on mood + quiz results.
5. App suggests what to study next.
