The Machine Learning and Data Science team builds a variety of both descriptive and prediction models, from the massive volumes of behavioral and description data generated by eBay’s many buyers, sellers, and products. The data at eBay is among the largest and most challenging/diverse in the world, with applications requiring the fusion of massive amounts of behavior log, text, and image data — putting us on the forefront of practical machine learning research.
We work closely with our partners in product groups, with a special focus on developing data-driven models which improve user experience, including product search. We are a collection of machine learning, systems engineering, and domain experts, who research and develop cutting-edge methods which scale to large eCommerce data involving hundreds of millions of active users and billions of transactions, and which leverage a variety of modern computing platforms, including both distributed (Hadoop) and massive-core (including GPU) computing. In addition to routinely targeting and shipping our work to improve the eBay site, our researchers actively engage external research communities in a variety of ways, including service, publication, and speaking at a variety of the top research conferences, including SIGIR, WWW, KDD, ICML, and NIPS.