Skip to content

Latest commit

 

History

History
4 lines (3 loc) · 674 Bytes

README.md

File metadata and controls

4 lines (3 loc) · 674 Bytes

binaryrecommender

The code of the online experiment that I conducted in order to perform A/B testing for my masters thesis. It consists of two recommender systems: one based on an algorithms developed for the thesis that incorporates pairwise preferences, the other based on matrix factorization - a state of the art in collaborative filtering. The website implemented using Play framework. Backend written in Java, frontend - in Javascript. DB used - PostgreSQL. Currently deployed on Heroku but the experiment is finished and it will be closed soon.