This project aims to develop a substrate called SocialLite that can use online social network data to obtain reliable identity and trust information. This work involves three steps: 1) identifying the rich variety of identity and trust information embedded in online social networks; 2) designing algorithms and software to efficiently and robustly abstract this information as a set of flexible API functions without violating a user?s privacy from large online social networks; and 3) evaluating the usefulness of the API by implementing a few sample applications. The SociaLite API can facilitate the development of robust and trustworthy distributed systems. Prototypes developed from this project will be released to the public to engage a broader developer community to use this API to build trustworthy applications. The algorithms developed in this project can advance the research community's understanding on trust metric computation and the utility of social-trust based computing. For further information, see the project website at www.cs.duke.edu/nds/wiki/sociallite.
