Online Product Recommendation Model based on Deep Learning

Modern e-commerce shopping web-sites with typically thousands of orders and billions of clicks from its large set of users rely on online product recommendation models. Each of these users has a buying history. Building an effective personalized recommendation remains still challenging due to the dynamic nature of the e-commerce. The aim of this project is to build an effective and responsive online recommendation model which benefits from users’ past purchase behaviour and product information.
Reciprocal Recommendation for Recruitment Web Sites: Reciprocal recommender models aim to satisfy the needs of both parties involved in the recommendation process. In this project a job recommendation model is built to match the qualifications of a candidate to the requirements of a position by considering the preferences of both sides. This is a joint project with an employment website from Turkey supported also by the Scientific Research Council of Turkey.

Sule Gunduz Oguducu
ITU Vision Lab
Funded by
April 2018