The Power of Known Peers: A Study in two Domains
by Prof. Peter Brusilovsky
School of Information Sciences
University of Pittsburgh
One of the frequently cited benefits of social recommender systems is the ability to use social connections that are explicitly represented in many social systems. While traditional community-driven personalization technologies such as collaborative filtering and social navigation use past behavior of nameless "peers" to guide the user to most appropriate items, social recommender systems can rely on friends and connections known to the user. However, can we demonstrate that recommendation approaches work better when fueled by "known peers" and provide a stronger impact on user behavior? In my talk I present two examples from two different domain - collaborative tagging systems and social e-learning systems that provide some new evidence that the use of "known peers" does change the recommendation game.
Speaker's short bio
Peter Brusilovsky is a Professor of Information Science and Intelligent Systems at the University of Pittsburgh, where he also directs Personalized Adaptive Web Systems (PAWS) lab. He has been working in the field of adaptive educational systems, user modeling, and intelligent user interfaces for more than 20 years. He published numerous papers and edited several books on adaptive hypermedia and the adaptive Web.
Picking the Best of the Best
by Shlomo Berkovsky
National ICT of Australia (NICTA)
Australian Technology Park, NSW, Australia
Social Networking sites often deploy the News Feed or Activity Feed as the means to keep their users abreast of the most important activities of other users. The feed in most cases shows the activities carried out by immediate friends of followees. However, as the social circles (and the feeds) of users continue to expand, the challenge of keeping tracks of truly important network activities invites more fine-grained solutions. In this talk we will overview several recent works into the personalisation of the feed, discuss their strong and weak sides, and outline future applications for the developed feed personalisation approaches.
Speaker's short bio
Shlomo Berkovsky is a senior researcher at the Next-Generation Content Discovery and Distribution project run by NICTA. The project is developing novel techniques for accurate personalisation and cost-effective delivery of catch-up and live TV services. Before joining NICTA, he was a senior research scientist and research team leader at CSIRO. Shlomo is interested in collaborative and content-based recommenders, mediation of user models, ubiquitous user modelling, context-aware personalisation, personalised persuasion, and privacy-enhanced recommendations. Shlomo is the author of more than 80 publications accepted to journals, books, and conference/workshop proceedings, and the chair of the upcoming Persuasive Technology 2013 conference.