This repo contains various algorithms we used for generating Top N recommendations
Data
The data sets have been used from Stanford Network Analysis group:(Julian McAuley, UCSD) and cannot be uploaded here as they were private and were provided on request.They contain real world anonymised customer data from Amazon webservices from 2014-2019.
Preprocessing/Analysis
Certain cleaning had to be done to remove the errors genrated while scraping the data from the websites, along with this user/item based analysis was done to get a good understanding of the data we would be using to make recommendations
Models
various models were implemented and results were analysed,the models being,
1)Content based Recommendations
2)Collaborative based Recommendations
3)Model based recommendations(Matrix factorizations)
4)Hybrid models
Please have a look at the project report for better understanding Report