AWS re:Invent 2019: Evolution of personalization/recommendation for video workflows
Personalizing the user experience is proven to increase discoverability, user engagement and satisfaction, and revenue; however, many AWS customers find personalization hard to get right. Effective recommender systems require solving multiple hard problems, including constantly changing user behavior, new catalog items (cold start), etc. In this session, we cover some of the most common methods of personalization/recommendation. Also, Amazon Prime Video shares the evolution of its recommendation system and the real-world challenges it faced when building recommendation systems at scale.