This is a Content-based recommendation system based on tmbd_5000_movies and tmbd_5000_credits datasets.
I used Streamlit to showcase my Recommendation System which some glimpses are given below.
Steps that have taken
Step 01: Initially I merged both the datasets on 'title', then took only those features which will facilitate me to apply Bag-of-Word(BOW).
step 02: After finishing the dataset, I did some preprocessing and feature extraction to draw clear data.
step 03: After cleaning the data, I applied text preprocessing(lower, stemming).
step 04: Applied CountVectorizer and then extracted similarity from cosine similarity.
step 05: Formulate the function by matching the index of the similarity matrix that gives 5 recommended movies based on the input movie name.
step 06: Dumping similarity data and new dataframe through pickle.
step 07: Load both the files in vscode and use them in streamlit syntax.
Author: Komal Prasad