Restaurant-Recommendation-Based-on-correlation

Purpose:

This Notebook Recommends Restaurants based on correlation

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 restaurants by different users.
2-We have a cuisine data which contains cuisines of different restaurants.
3-We have gio locations data which contains locations and names of restaurants.

Libraries Used:

1-Pandas
2-Numpy

Approach

1-From the ratings data we group the data by restaurant's ID.
2-Then we find the count of ratings for each restaurant and averge rating.
3-Hence we will get a data frame with restaurants, their count of ratings and average rating.
4-Merge this data frame with most rated restaurants.
5-Now find a pivot table for each users rating of each restaurant.
6-Find the most rated restaurant and find its correlation with other restaurants.
7-Add the correlation with ratings data.
Hence we can recommend a Restaurant based on high correlated with most rated restaurant and its cuisine.