Personalisation and Recommender Systems

Personalization is one of the great promises of AI. Personalization means that solutions to problems in areas such as medicine, education, route planning and many more can be tailored automatically to suit our individual preferences and constraints. Personalization lies at the heart of Recommender Systems, which are software systems that help us to manage choice. Faced with an overwhelming volume of products, services and sources of information, a recommender system ranks the choices according to what it has learned about its users’ preferences and about the context. Recommender Systems are among the most successful and wide-scale uses of AI: we find them in movie & music streaming services, online shopping sites, and social media platforms. But, personalization and recommender systems done badly can invade privacy and promote negative content. The research challenges in personalization and recommender systems might be summarized as how to enjoy the benefits without being negatively impacted from the downsides. Our research in the Centre seeks to address these challenges.