Putting the Role of Personalization into Context

Abstract

Personalization is omnipresent in our life, with applications ranging from entertainment and commercial uses to smart devices and medical treatments. The integration of personalization in various products turned rapidly from an unnecessary luxury to a commodity that is expected by customers. While different machine learning fields present state-of-the-art advances and super-human performance, personalization applications are often late-adopters of novel solutions due to their complex framing and multiple stakeholders' with different business goals. The role of personalisation applications is also ambiguous: it is unclear, for instance, whether models just predict a user's next action or proactively affect the user's selections. This talk focuses on examining the role of recommenders and their ability to adapt to customer feedback. Key topics such as causality and active exploration are depicted with real examples and demonstrated alongside business considerations and implementation challenges. It relies on recent advances in the field and on work conducted at Booking.com, where we implement personalization models on one of the world's leading online travel platform.