🔹 Introduction This project is a modern web-based platform designed for student onboarding, course recommendations, and peer-pairing using machine learning. Built using the Vite framework, it provides a fast, modular development experience. The application includes user authentication, profile management, interest-based course suggestions, and an intelligent student pairing system powered by K-Means Clustering. The long-term vision includes integrating a personalized AI chatbot for academic support and mentorship.
🔹 Tech Stack
| Layer | Technology |
|---|---|
| Frontend Framework | Vite + React |
| Styling | Tailwind CSS / CSS Modules |
| Routing | React Router DOM |
| State Management | React Context / useState (or optionally Redux) |
| Machine Learning | K-Means Clustering (Python/JavaScript) |
| Backend (if used) | Node.js + Express |
| Database | MongoDB |
| Authentication | Custom Auth API |
| AI Integration | Custom LLM-based chatbot (e.g. OpenAI API, Cloudflare Workers AI) |
🔹 Features
✅ Authentication Pages
- Register Page: Allows new users to sign up with personal and academic info.
- Login Page: Secure login system for registered users.
✅ Main Menu Page
Contains navigational links for:
- Edit Profile: Users can update personal details, interests, or preferences.
- Get Courses: Dynamically displays courses based on interest or cluster group.
- Add Interest: Users can tag areas of interest to improve course suggestions and clustering accuracy.
🔹 Student Clustering Logic (K-Means)
🎯 Goal: To group students based on knowledge scores, interests, or other metrics into clusters for more meaningful interactions.
⚙️ How It Works:
- K-Means Clustering algorithm is used to divide the student population into fixed-size groups (e.g., 25 students per cluster).
- Clustering inputs may include: knowledge_score, selected interests, previous course completions, etc.
- Each student is assigned to a
cluster_id, saved in the backend.
🤝 Pairing Logic:
- After clustering, students are paired within the same cluster based on complementary strengths (e.g., high scorer paired with a mid-level one).
- This pairing is intended to promote collaborative learning and peer mentoring.
🧠 Future Plans: AI Chatbot Integration We plan to introduce a personalized AI chatbot that:
- Provides course recommendations
- Explains difficult topics
- Suggests learning paths
- Gives tailored feedback based on progress and interests
Possible frameworks:
- OpenAI GPT-based bot
- Custom LLM hosted on Cloudflare Workers AI
- Integration with course data and student clustering for personalization