A content-based movie recommendation system with a sleek, Netflix-inspired user interface.

[Github Page App Link]

CineMatch is a movie recommender system that uses content-based filtering to suggest movies based on user preferences. The system analyzes movie metadata including genres, keywords, cast, crew, and plot overviews to find similar movies.

• Netflix-Style UI: Dark theme with smooth animations and modern design
• Smart Recommendations: Content-based filtering using cosine similarity
• Movie Posters: Integration with TMDB API for dynamic poster display
• Match Scores: Visual similarity percentages for each recommendation
• Responsive Design: Works on desktop and mobile devices

The system uses the TMDB 5000 Movie Dataset which contains:

• Movie metadata (genres, keywords, overview)
• Cast and crew information
• Production details

Each movie is represented by a "tag" combining:

• Overview text (stemmed)
• Genre names
• Top 3 cast members
• Director name
• Keywords

from sklearn.feature_extraction.text import CountVectorizer

cv = CountVectorizer(max_features=5000, stop_words='english')
vectors = cv.fit_transform(movies['tags'])

from sklearn.metrics.pairwise import cosine_similarity

similarity = cosine_similarity(vectors)

def recommend(movie):
    movie_index = movies[movies['title'] == movie].index[0]
    distances = similarity[movie_index]
    movies_list = sorted(list(enumerate(distances)), 
                         reverse=True, key=lambda x:x[1])[1:6]
    return [movies.iloc[i[0]].title for i in movies_list]

1. Heroku Account
2. Heroku CLI
3. Git
4. TMDB API Key (optional, for posters)

# 1. Clone or create your repository
git init movie-recommender
cd movie-recommender

# 2. Add all files (app.py, requirements.txt, Procfile, setup.sh, pickle files)

# 3. Login to Heroku
heroku login

# 4. Create a new Heroku app
heroku create your-app-name

# 5. Set config variables (optional, for movie posters)
heroku config:set TMDB_API_KEY=your_api_key_here

# 6. Deploy
git add .
git commit -m "Initial deployment"
git push heroku main

# 7. Open your app
heroku open

movie-recommender/
├── app.py                 # Main Streamlit application
├── requirements.txt       # Python dependencies
├── Procfile              # Heroku process file
├── setup.sh              # Streamlit configuration
├── runtime.txt           # Python version
├── movies_dict.pkl       # Movie data (serialized)
├── similarity.pkl        # Similarity matrix (serialized)
└── README.md             # Documentation

1. Select a Movie: Use the dropdown to search for a movie you've enjoyed
2. Get Recommendations: Click the "Get Recommendations" button
3. Explore Results: View 10 similar movies with match percentages
4. Discover More: Click on movies to learn more about them

• Collaborative filtering integration
• User authentication and watch history
• Movie trailers and streaming links
• Genre-based filtering
• Multi-language support

Jay Dayal Guwalani UID: 121479709 MSML602 - Principles of Data Science
University of Maryland

• TMDB 5000 Movie Dataset
• Streamlit Documentation
• Heroku Python Deployment
• Cosine Similarity Explained

This project is for educational purposes as part of MSML602 coursework.