Recommendation System for E-Commerece Platform

Data Preparation
The dataset is loaded from the Electronics_5.json file.
Only the first 5000 lines are read to keep the dataset manageable.
A pandas DataFrame is created with relevant columns: reviewerID, asin, and overall.
Users who have rated 5 or more items are retained for analysis.
A pivot table is created where rows represent users and columns represent products, with ratings as values.

Exploratory Data Analysis
Summary statistics of the ratings, number of unique users and products, and density of the ratings matrix are calculated.
The top 10 users based on the number of ratings are identified.

Popularity-Based Recommendation
Products are recommended based on the count of reviews in the training data.
A function to recommend the top 5 products to any user based on popularity.
The user-item ratings matrix is factorized using SVD to reduce dimensionality and make predictions.
A function to recommend items to a user based on the predicted ratings.