Peer-to-Peer Data Sharing for Integration and Collaboration

Zack Ives
University of Pennsylvania

A major challenge lies in effective sharing of information among
collaborating, yet autonomous, parties.  These parties might be
institutions, users, or even sensor devices (or combinations of sensor
devices).  They are characterized by having a diversity of perspectives (and
hence heterogeneous schemas), dynamic data, and the possibility of
intermittent connectivity or participation.  The parties are peers in the
sense that they are fully autonomous, they contribute and use resources as
they choose, and they may join or leave at any point.

In this talk, I will provide an overview of our recent efforts to develop
rapid, extensible, and flexible data sharing capabilities, using a
peer-to-peer-inspired approach, in three different settings:  SHARQ, which
focuses on rapid sharing and querying of data; Orchestra, which addresses
the challenges of propagating changes across interrelated databases; and
Aspen, which focuses on sensor data integration, actuation, and management
of confidentiality.

Joint work with Val Tannen, Susan Davidson, Insup Lee, Nick Taylor, TJ
Green, Grigoris Karvounarakis, and Madhukar Anand.