🎬 AI Movie Recommendation System

An intelligent movie recommendation system built using Machine Learning + FastAPI + Streamlit. This project suggests similar movies based on semantic similarity using AI embeddings.

---

🚀 Live Demo

• 🌐 Streamlit App: (https://ai-recommendation-system-streamlit.onrender.com)
• ⚡ FastAPI Backend: (https://ai-recommendation-system-fastapi.onrender.com)

⚠️ Important: Please run the FastAPI backend first before using the Streamlit app.

Otherwise, the app will not return any responses.

---

🚀 Features

• 🔍 Smart movie search with auto-suggestions
• 🤖 AI-based recommendations using sentence embeddings
• 🎯 Fuzzy search (handles typos like "interstella" → "interstellar")
• 🖼️ Movie posters fetched from TMDB API
• ⭐ Similarity score for each recommendation
• 🎬 Clean UI with Streamlit (grid layout)
• ⚡ FastAPI backend for high performance

---

🧠 How It Works

1. Movie data is loaded from a dataset (dataset.csv)

2. Text features are combined:

   • Synopsis
   • Genre
   • Director
   • Content

3. AI embeddings are generated using:

   • sentence-transformers (all-MiniLM-L6-v2)

4. Cosine similarity is used to find similar movies

5. Results are filtered (genre + quality filters)

6. Posters are fetched from TMDB API

---

🏗️ Tech Stack

• Python
• FastAPI (Backend API)
• Streamlit (Frontend UI)
• Scikit-learn (Cosine similarity)
• Sentence Transformers (Embeddings)
• RapidFuzz (Fuzzy search)
• TMDB API (Movie posters)

---

📁 Project Structure

AI-Movie-Recommender/
│
├── api/
│   └── main.py              # FastAPI backend
│
├── model/
│   └── recommender.py      # Recommendation logic
│
├── data/
│   └── dataset.csv         # Movie dataset
│
├── dashboard/                  # Streamlit frontend
│    └──app.py
├── embeddings.npy          # Precomputed embeddings
└── README.md

---

⚙️ Installation

1. Clone Repository

git clone https://github.com/wpko/AI-Recommendation-System.git
cd AI-Recommendation-System

---

2. Create Virtual Environment

python -m venv venv
venv\Scripts\activate   # Windows

---

3. Install Dependencies

pip install -r requirements.txt

---

🔑 TMDB API Setup

1. Create account at: https://www.themoviedb.org/
2. Get your API key
3. Add it in main.py:

API_KEY = "your_tmdb_api_key"

---

▶️ Run the Project

Step 1: Start FastAPI Backend

python -m uvicorn api.main:app --reload

👉 API will run at: http://127.0.0.1:8000

---

Step 2: Start Streamlit Frontend

streamlit run app.py

👉 UI will open in browser automatically

---

🧪 Example Usage

• Input: interstella

• Output:

  • Interstellar-like movies
  • Posters + similarity scores

---

📸 Screenshots

Streamlit Screenshots

FastAPI Screenshot

---

🎯 Future Improvements

• 🔥 Personalized recommendations
• 🎞️ Trailer integration (YouTube API)

---

🤝 Contributing

Feel free to fork this repository and improve it!

---

👨‍💻 Author

Wai Lay Aspiring AI / Python Developer 🚀

---

⭐ If you like this project, give it a star on GitHub!