Collective intelligence has been harnessed recently to create new collaborative forms that were not previously achievable. This socially-emergent intelligence appears in collaborative management and indexing of information, authoring encyclopedia articles, participating in blogs, tagging or commenting photographs or videos. This socially-emergent movement also exposes an unprecedented amount of social information. Particularly, two classes of Internet applications reveal voluminous social information: popular online social networks (e.g., Facebook or LinkedIn); and the widely-adopted collaborative tools (e.g., CiteULike or Delicious) that provide rich information fabric through tags, annotations, and text organization.
This project investigates the potential of including social knowledge in the design of community-enabled peer-to-peer distributed infrastructures. It will design, prototype, and evaluate a community-oriented peer-to-peer infrastructure that exploits social knowledge for services such as data and computing management, while protecting social data privacy seen as contextual integrity. Using social knowledge in the design of services will likely improve performance, based on the assumptions that social incentives reduce churn; socially-inferred trust reduce security breaches and expand the set of available resources; and shared interest improves data placement decisions or co-location of data and computations. This will create the potential to enable new classes of applications and infrastructures.
This project develops innovative teaching strategies that facilitate non CS-major students use of computational tools in their professional lives. This will be done through multidisciplinary co-teaching of an undergraduate introductory course, and coordinated teaching of independent graduate courses in computer science and social sciences.
