Real-Time Event Recommendationby Cliff Lyon
Head of Technology Recommender Systems
Personalized content keeps customers interested, encourages loyalty, and drives sales. As the recommendation technology group, our challenge is to build a flexible and powerful recommender system to support personalization, which allows for deep targeting, rapid response to behavior changes, and easy feature development. In order to move beyond our first version recommendation engine, we developed a way to construct on-the-fly user profiles, allowing us to recommend in real time, as well as to develop new features and methods of recommendation easily. In this talk, I will talk about our system, the data flow, and the challenges and lessons learned so far.