Overview The Course Recommendation System is a machine learning-powered web app built with Streamlit that suggests courses based on similarity of course embeddings. It leverages BERT-based embeddings and cosine similarity to find the best matching courses.

Features

• 🔍 Course Selection: Choose a course from the dropdown, and get top 5 recommended courses based on similarity.
• 📌 Recommendation Cards: Displayed in aesthetic semi-transparent cards with course details.
• 🏫 University & Ratings: Each recommendation includes university name and course rating.
• 🌐 Course Search: Search for courses by name.
• 💾 Processed Data Export: Save the processed courses as a CSV file.

Screenshots

Installation & Usage

git clone https://github.com/Ad-Chekk/CRS.git
cd CRS

2️⃣ Install Dependencies

pip install -r requirements.txt

3️⃣ Run the Application

streamlit run app.py

📂 Project Structure

📁 course-recommendation/
├── 📄 app.py            # Streamlit UI & Recommendation Logic
├── 📄 course_embeddings.pkl # Precomputed Course Embeddings
├── 📄 processed_embeddings.csv # Expanded Embeddings Data
├── 📄 processed_courses.csv # Processed Course Data
├── 📄 requirements.txt  # Dependencies
└── 📄 README.md         # Project Documentation

How It Works

1. Loads Precomputed Embeddings: Reads course_embeddings.pkl (created using BERT model).
2. Processes Embeddings: Converts tensors to lists and expands them into individual dimensions.
3. Computes Similarity: Uses cosine similarity to find the closest matching courses.
4. Displays Recommendations: Top 5 courses are shown in beautifully styled semi-transparent cards.

🤖 Technologies Used

• Python 🐍
• Streamlit 🎨 (For UI)
• Torch (PyTorch) 🔥 (For Tensor Handling)
• scikit-learn 🧠 (For Similarity Computation)
• pandas 🏗 (For Data Processing)

📌 Future Enhancements

• ✨ User-based personalization (Store user preferences for better recommendations)
• 🔥 More advanced NLP models (Try sentence transformers for better embeddings)
• 🌍 Deploy on a cloud platform (Streamlit Cloud / AWS / Heroku)

Contributions are welcome! Feel free to fork the repo and submit a pull request.

This project is licensed under the MIT License.