🎮 CineFinder

A semantic search engine to help users discover movies and TV shows using natural language queries. CineFinder uses vector embeddings and FAISS indexing to provide fast, intelligent, and context-aware recommendations.

---

📌 Features

• 🔎 Semantic Search with vector similarity using FAISS
• 🧠 Embeddings generated from movie titles and descriptions
• ⚡ Fast retrieval from a dataset of 6K+ movie entries
• 📊 Interactive Jupyter Notebook for exploration and testing

---

🧱 Project Structure

CineFinder/
├── Movie_Wiki_Dataset.ipynb    # Notebook to clean data, embed, and search
├── movie_data.pkl              # Pickled movie metadata
├── movie_faiss.index           # FAISS vector index for fast search
├── requirements.txt            # Python dependencies

---

🚀 Getting Started

1. Clone the Repository

git clone https://github.com/ShouryaAswal/CineFinder.git
cd CineFinder

2. Install Dependencies

pip install -r requirements.txt

3. Run the Notebook

jupyter notebook Movie_Wiki_Dataset.ipynb

💡 Use the notebook to embed text, build the FAISS index, and run semantic queries like:

search("time travel thriller with a twist ending")

---

🔧 Tech Stack

• Python & Jupyter Notebook
• FAISS for vector similarity search
• Pickle for data serialization
• Scikit-learn / Transformers (optional for embeddings)

---

📬 Contact

For questions or collaborations:

• 📧 shouryaaswal@hotmail.com
• 🌐 github.com/ShouryaAswal

---

🎥 CineFinder is an experimental semantic search engine — perfect for exploring how AI can enhance content discovery in entertainment platforms.