PEPPER: Peer-to-Peer Data Management Presented by: Frans Adjie Effendi for

Peer-to-peer systems have made inroads as scalable content distribution networks. Most current systems allow for the retrieval of a document given the name of the document, although recent work has addressed somewhat more expressive queries based on keyword searches.

As part of the PEPPER project, we are exploring a much richer query semantics for peer-to-peer systems. We envision a future where users will publish data at their local servers as annotated XML document, whose DTDs conform to some industry or societal standards. A highly scalable, distributed data management infrastructure will allow users to query this web data warehouse as if the data were stored in one huge centralized database system. Towards this goal, we are developing novel peer-to-peer indexing and query processing techniques.

Publications

A. Crainiceanu, P. Linga, J. Gehrke, J. Shanmugasundaram, "Query Peer-to-Peer Networks Using P-Trees", International Workshop on Web and Databases (WebDB), June 2004.

A. Crainiceanu, P. Linga, A. Machanavajjhala, J. Gehrke, J. Shanmugasundaram, "A Storage and Indexing Framework for P2P Systems", WWW Conference (poster), May 2004.

Demonstrations

A. Crainiceanu, P. Linga, A. Machanavajjhala, J. Gehrke, J. Shanmugasundaram, "An Indexing Framework for P2P Systems", SIGMOD Conference, June 2004.

Technical Reports

A. Crainiceanu, P. Linga, A. Machanavajjhala, J. Gehrke, J. Shanmugasundaram, "P-Ring: An Index Structure for Peer-to-Peer Systems", Cornell University Technical Report, July 2004.

People

Adina Crainiceanu
Frans Effendi
Johannes Gehrke
Nicolas Hamatake
Prakash Linga
Linda Nguyen
Jayavel Shanmugasundaram

Alumni

Guillaume Giraud
Ashwin Machanavajjhala
Tat-Yeong Cheah

Acknowledgment

This project is funded in part by the National Science Foundation and a gift from Microsoft Corporation.