People often rely on the collective intelligence of
their social network for making choices, which in turn influences
their preferences and decisions. By treating recommendations as embedded in a social network, many interesting questions emerge. Network-centric
recommendations provide a domain that invites looking
at the bigger picture of how recommendations interact with the world.
How might accounting for influence affect algorithm design?How does interface design affect influence processes? And how do the recommendations generated by those algorithms in turn affect the people and the networks that receive them?
Our first work in this space is PopCore, a project that aims to use your friends' interests and preferences to better understand your tastes and provide recommendations. Currently, we are looking at the entertainment domain, including books, movies and tv shows. It runs as an app on Facebook, and soon you'll be able to try it out.
Active: (Faculty) Dan Cosley (Phd) Amit Sharma
Past: (Masters) Sneha Kanneganti, Michael Triche, Shruti Gautam (Undergraduate) Yulan (Lannie) Miao, David Bodin, David Bodin (Staff) Meethu Malu
- Sharma, A., Gemici, M., Cosley, D. (2013). Friends, Strangers, and the Value of Ego Networks for Recommendation. ICWSM 2013 poster paper. [PDF] [Publisher]
- Sharma, A., Cosley, D. (2013). Do Social Explanations Work? Studying and Modeling the Effects of Social Explanations in Recommender Systems. WWW 2013. 1133-1144. [PDF] [Publisher]
- Sharma, A., Cosley, D. (2011). Network-Centric Recommendation: Personalization with and in Social Networks. SocialCom 2011. 282-289. [PDF] [Publisher]
- Sharma, A., Malu, M., Cosley, D. (2011). PopCore: A system for Network-Centric Recommendations. 3rd Workshop on Recommender Systems and the Social Web. [PDF]