Skip to content
/ ozilly Public

Ozilly is a Netflix-style movie recommendation system built with Streamlit that provides personalized suggestions using cosine similarity, featuring genre filtering, fuzzy search, and a clean, minimalist user interface.

Notifications You must be signed in to change notification settings

yaadhuu/ozilly

Folders and files

NameName
Last commit message
Last commit date

Latest commit

Β 

History

7 Commits
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 
Β 

Repository files navigation

🌌 Ozilly – A Curated Cinematic Companion

Ozilly is not just a movie recommendation engine β€” it's a refined, curated, and personalized cinematic discovery experience. Built with precision, served with elegance.


✨ Why Ozilly?

  • 🎞️ Curated Recommendations β€” Discover films that match your cinematic fingerprint.
  • 🧠 Smart Matching β€” Suggests movies using cosine similarity, not generic trends.
  • πŸ” Fuzzy Title Search β€” Handles typos and partial entries with intelligence.
  • πŸ—‚οΈ Genre Filtering β€” Browse by theme, mood, or aesthetic.
  • πŸ•’ Recent History Tracking β€” Pick up where you left off, effortlessly.
  • πŸ’– Favorites Manager β€” Save what you love in a click.

πŸ”§ Tech Stack

Layer Tech Used
Backend Python, Pandas, Pickle
Recommender Scikit-Learn (cosine sim)
UI Engine Streamlit (custom styled)
Assets TMDB Posters API (optional)

🧠 How It Works

  1. Loads movies from movie_dict.pkl (preprocessed metadata)
  2. Uses a similarity.pkl matrix built on cosine similarity
  3. Based on selected movie, returns top 5 related titles
  4. Displays clean poster cards with refined layout

πŸš€ Run Ozilly Locally

git clone https://github.com/yaadhuu/ozilly.git
cd ozilly
pip install -r requirements.txt
streamlit run app.py

About

Ozilly is a Netflix-style movie recommendation system built with Streamlit that provides personalized suggestions using cosine similarity, featuring genre filtering, fuzzy search, and a clean, minimalist user interface.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors