This project implements variations of Bayesian Personalised Ranking and compare the perfomances of each with otehr. We tried to understand if users' social connections can be leveraged to improve rankings of items recommended to them by a recommender system. We implemented the below mentioned paper and refered to github repositories to get started with our project.
References (cloned in the references directory):
-
Leveraging Social Connections to Improve Personalized Ranking for Collaborative Filtering
-
bpr by
gamboviol
-
BPR_MPR by
RunlongYu
NOTE: References to be added here
- When you want to add another git repo in yours (cloned), add it using the following git command:
git submodule add <repo-link> dir/repo-name