PoolView
Authors:
Open Access
- 5 November 2008
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 281-294
Abstract
This paper develops mathematical foundations and architectural components for providing privacy guarantees on stream data in grassroots participatory sensing applications, where groups of participants use privately-owned sensors to collectively measure aggregate phenomena of mutual interest. Grassroots applications refer to those initiated by members of the community themselves as opposed to by some governing or official entities. The potential lack of a hierarchical trust structure in such applications makes it harder to enforce privacy. To address this problem, we develop a privacy-preserving architecture, called PoolView, that relies on data perturbation on the client-side to ensure individuals' privacy and uses community-wide reconstruction techniques to compute the aggregate information of interest. PoolView allows arbitrary parties to start new services, called pools, to compute new types of aggregate information for their clients. Both the client-side and server-side components of PoolView are implemented and available for download, including the data perturbation and reconstruction components. Two simple sensing services are developed for illustration; one computes traffic statistics from subscriber GPS data and the other computes weight statistics for a particular diet. Evaluation, using actual data traces collected by the authors, demonstrates the privacy-preserving aggregation functionality in PoolView.Keywords
This publication has 24 references indexed in Scilit:
- The Dynamics of Human Body Weight ChangePLOS Computational Biology, 2008
- The BikeNet mobile sensing system for cyclist experience mapping2007
- Image browsing, processing, and clustering for participatory sensing2007
- An ODE traffic network modelJournal of Computational and Applied Mathematics, 2007
- CenWits2005
- Achieving Privacy Preservation when Sharing Data for Clustering2004
- Randomization in privacy preserving data miningACM SIGKDD Explorations Newsletter, 2002
- Classroom Note:A Model of DietingSIAM Review, 1998
- General choice of the regularization functional in regularized image restorationIEEE Transactions on Image Processing, 1995
- Randomized Response: A Survey Technique for Eliminating Evasive Answer BiasJournal of the American Statistical Association, 1965