🎞️ Movie Finder 🎞️

A intelligent movie recommendation system that uses advanced techniques to recommend movies based on your preferences.

✨ Key Features

• Platform-independent movie recommendations
• Genre Selection
• Mood Choice
• Era Timeframe Options
• Language Decision
• Notes for in-depth personalization

🚀 Getting Started

💻 Prerequisites

• Git
• Bun.js (OR: npm & node)
• Python 3.11+

📦 Installation

# Clone the repository
git clone git@github.com:IImpaq/movie-finder.git
cd movie-finder

# Prepare the backend
cd backend
pip install -r requirements.txt
python fetch.py      # Fetch dataset from huggingface
python preprocess.py # Prepare and preprocess the data
cd ..

# Prepare the frontend
cd frontend
bun install # OR: npm install
cd ..

⚙️ Running

Full-Stack Application

# Terminal 1: Start Frontend
cd frontend
bun run dev # Website Available at http://localhost:3000

# Terminal 2: Start Backend
cd backend
uvicorn app:app --reload  # API available at http://localhost:8000

CLI Mode

# Run backend/movie recommender in interactive cli mode
cd backend
python cli.py

Evaluate the recommender system

# Run the automated evaluation script
cd backend
python evaluation.py

📚 Dataset

Already included is the preprocessed dataset. It includes around 180.000 rows and the following columns:

["id", "title", "genres", "original_language", "overview", "popularity", "vote_average", "release_date", "status", "keywords", "credits", "poster_path"]

The preprocessed dataset is generated from the raw data of the wykonos/movies collection that is published on Hugging Face.

🔧 Tech Stack

• Frontend: NextJS & TypeScript
• Backend: Python, FastAPI & PyTorch

🔗 Links

• Design Document
• Report
• Short Presentation
• Full Presentation
• Questionnaire
• Evaluation
• Movie Dataset

🧠 Team & Roles

• Patrick Eckel: Design Document, Frontend, Data Transfer
• Marcus Gugacs: Design Document, Frontend, Recommender, CLI, Evaluation, Questionnaire, Report, Presentation
• Martin Tobias Klug: Design Document, Subtitle Fetching, Summarization Pipeline, Report
• Lukas Leitner: Design Document, Data Preprocessing, Report, Presentation

📝 License

MIT License (see LICENSE).

☎️ Contact

If you have any questions or want to get in touch, just send an email

---

Made with ❤️