Title: Two Brains are Better than One: User Control in Adaptive Information Access
Speaker: Peter Brusilovsky
Summary and thoughts:
Dr. Peter Brusilovsky is a renowned expert in recommender systems for education, health, and social recommendations. His talk at the latest ISP forum focused on the ability and use of human and artificial brains to work together to implement adaptive information access systems. His projects over the last few decades have focused on helping people to get to the right information at the right time through systems designed to search, navigate, and recommend. His work in the field of human-centered AI aims to implement user control by engaging the end users of AI and ones affected by the AI’s decisions in development and design. The talk discussed several projects focused on user control in adaptive information access systems including adaptive annotation for learning content (1996), adaptive hypermedia search, controllable personalized search (TaskSieve), exploratory search for scientific purpose (SciNet), conference navigator, and a social recommender system (RelevanceTuner). The specific themes involved in user control in adapative information access systems include exploration, control, and transparency. A separate section discussed the importance of visualization in recommendation systems to allow users to explore the system and help them understand the underlying mechanisms. A common, recurring theme throughout the projects displayed was the active role of users in evaluation of the performance and design of the systems. Dr. Brusilovsky’s work reaffirms that combining the decision power or AI with the ability of the user to guide and control it combines the strong side of artificial and human intelligence and leads to better results.