Plot description-based recommender

Our plot description-based recommender will take in a movie title as an argument and output a list of movies that are most similar based on their plots. These are the steps we are going to perform in building this model:

1. Obtain the data required to build the model
2. Create TF-IDF vectors for the plot description (or overview) of every movie
3. Compute the pairwise cosine similarity score of every movie
4. Write the recommender function that takes in a movie title as an argument and outputs movies most similar to it based on the plot

Preparing the data

In its present form, the DataFrame, although clean, does not contain the features that are required to build the plot description-based recommender. Fortunately, these requisite features are available in the original metadata file.

All we have to do is import them and add them to our DataFrame:

#Import the original file
orig_df = pd.read_csv('../data/movies_metadata.csv', low_memory=False)

#Add the useful features into the...