Content-based Movie Recommender Application

Try out the Movie-Recommender Application. (Deployed using streamlit)

This Repository contains a Machine Learning Project. It is a content-based movie recommender system which uses the tmdb 5000 movies and credits kaggle dataset to recommend 5 movies based on a movie watched.

Some of the main points to mention about this project are as follows:

• It is a content-based system of recommendation.
• The tmdb 5000 movies and credits datasets have been used: Kaggle Dataset Link
• It utilizes the technique of Text Vectorization using tags from the movie overview, crew details, genre etc.
• There are various ways to vectorize text, however, I have chosen the simple but effective technique of Bag of Words .
• Some of the major python modules used are numpy, pandas, nltk, pickle, tmdbv3api, requests and streamlit.
• Cosine Distance has been used as the criteria for determining similarity (closeness) between the vectors. (In higher dimensional spaces, Euclidean distance is not a differentiating factor)

Steps to run and execute the application locally:

To execute the file, simply clone the repository in your local machine, download the datasets from the Drive Link and execute the entire .ipynb file. Once it is done, you will have two files similarity.pkl and movie_dictionary.pkl. After this, simply open the terminal and execute app.py.

streamlit run app.py

OR, if you wish to save time, you can simply download the .pkl files too from the given Drive Link.

P.S. Ensure that all the files are in the same directory before running app.py and follow the steps carefully.