Time-aware Collaborative Filtering for eBay Auctions

Recommender systems are in wide use today by major companies (e.g. Amazon, Netflix and Google) to automatically direct their users to content that they would be interested without the tedium of sifting through the large selection of content available today.  This project is interested in performing an analysis and implementation of a recommender system for the popular auction site eBay.  The system should be capable of taking into consideration both what to recommend and when to recommend it; it will also need to be highly scalable as we will be using a large dataset from Advanced E-commerce Research Systems (AERS).  When completed, the project should provide AERS with at least a proof-of-concept for a recommender system that could be used to develop an operational system for eBay recommendations as well as provide analytics for eBay researchers.

Intern: 
John Chia
Faculty Supervisor: 
Dr. Alan Wagner and Dr. Nando de Freitas
Province: 
British Columbia
Discipline: 
Program: