In the present project, I create a content-based recommender system for American movies up to the year 2020. The project consists of the following steps.

The movies up to 2017 was taken from here and here. For the years from 2018 to 2020, I used the Wikipedia page for the American movies and used the BeautifulSoup to scrape the tables. Click this for the 2020 movies. Some additional data like genres were then obtained with the help of TMDb API.

To classify the reviews for the movies I used TF-IDF and the Random Forest Classifier which gave a 98.8% accuracy.

To recommend the top 10 movies, I used a simple, yet very powerful cosine similarity algorithm.