A Scalable Recommender System for one of Europe's largest e-commerce Platforms
Congratulations to Barrie Kersbergen, who defended his PhD thesis on 'Expanding boundaries in scalable session-based recommendations' at the University of Amsterdam. Barrie is an external PhD student, co-supervised by BIFOLD Professor Sebastian Schelter from the DEEM Lab.
Barrie has been doing a PhD next to his job as a staff scientist at Bol.com, a large European e-commerce platform in the Netherlands with more than 13 million active customers. In the last four years, Barrie and Sebastian have conducted research on improving the scalability, response latency, deployment cost and ecological sustainability of large-scale recommender systems. A special feature of this PhD collaboration is that the research results are already applied in the real world! Every second, thousands of recommendations on the product pages of bol.com are generated for Bol's customers using the recommender system 'Serenade' from Barrie's PhD research.
This collaboration serves as a prime example for impactful research at the intersection of data engineering and machine learning with a short path to the application in practice, and a ready-to-deploy open source software as outcome.
Publication: Expanding boundaries in scalable session-based recommendations. Barrie Kersbergen. Superviser: Maarten de Rijke, Sebastian Schelter. [Link]