AzeemWaqarRao / Model-Based-Movie-Recommendation-System

This Notebook Recommends Movies by finding correlation based on user rating of each movie

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Model-Based-Movie-Recommendation-System

Purpose:

This Notebook Recommends Movies by finding correlation based on user rating of each movie

Correlation:

Customer A likes restaurants 1 and 2.
Customer B like restaurant 2 so he's most probable to like restaurant 1 too.

Data:

1-We have a ratings data which contains ratings given to different movies by different users.
2-We have a movies data which contains names and IDs of movies.

Libraries Used:

1-Pandas
2-Numpy
3-SKlearn

Approach

1-The Data contains rating of different movies given by different Users.
2-Then we will group the data by movie IDs and find the count of ratings received by each movie
3-Then we will create a pivot table.
4-It will take users as rows and movies as columns
5-The cell will be filled by the rating they give to a movie they watched
6-Null values will be filled by 0
7-There are 943 users
8-we will compress the data to have 1664 * 943 to 1664*12
9-Then we will find correlation of every movie with each other based on user ID
10-If a person rated starwars high, he's likely to enjoy the movies that have high correlation with starwars.

About

This Notebook Recommends Movies by finding correlation based on user rating of each movie


Languages

Language:Jupyter Notebook 97.9%Language:Perl 1.1%Language:Shell 1.0%