This is a content-based movie recommender system built using Python and Streamlit. It suggests movies similar to a user-provided title by analyzing metadata such as genre, keywords, cast, director, and tagline using TF-IDF and cosine similarity.
Access the app here:
https://movierecommender-divyeb.streamlit.app
- Enter your favorite movie name in a search box.
- The system finds the closest matching movie title from the dataset.
- Top 30 similar movies are recommended based on content similarity.
- Built using a simple and clean Streamlit interface.
- Missing values in the metadata are filled.
- Features like genres, keywords, cast, director, and tagline are combined into a single string.
- The combined text is vectorized using TF-IDF.
- Cosine similarity is used to measure how close movies are to each other.
- The app uses fuzzy matching to handle slight spelling errors or typos.
The recommendation system uses a movie metadata file (movies.csv) that includes:
- Movie titles
- Genres
- Keywords
- Taglines
- Cast & Directors
This data is used only within the deployed app and is not intended for public reuse.
Divye Bajaj
LinkedIn