It is a content-based recommendation system. A content-based recommender works on the data we collect from the user. This recommender works by taking a movie title as an input and recommends 5 similar movies to it.
I used bag of words (CountVectorizer) for text vectorization and cosine distance to determine the similarity of vectors.
- The movies records are considered as vectors.
- Using cosine distance, closest vectors are compared.
- The closest 5 vectors to the user input are displayed as an output to the user.
The web interface is created using streamlit, an open-source app framework in Python.
https://www.kaggle.com/datasets/tmdb/tmdb-movie-metadata
https://movies-recommendation-system.streamlit.app/
streamlit run app.py