DocumentCode :
3635258
Title :
Anonymizing weighted social network graphs
Author :
Sudipto Das;Ömer Eğecioğlu;Amr El Abbadi
Author_Institution :
Department of Computer Science, University of California, Santa Barbara, 93106-5110, USA
fYear :
2010
fDate :
3/1/2010 12:00:00 AM
Firstpage :
904
Lastpage :
907
Abstract :
The increasing popularity of social networks has initiated a fertile research area in information extraction and data mining. Although such analysis can facilitate better understanding of sociological, behavioral, and other interesting phenomena, there is a growing concern about personal privacy being breached, thereby requiring effective anonymization techniques. In this paper, we consider edge weight anonymization in social graphs. Our approach builds a linear programming (LP) model which preserves properties of the graph that are expressible as linear functions of the edge weights. Such properties form the foundations of many important graph-theoretic algorithms such as shortest paths, k-nearest neighbors, minimum spanning tree, etc. Off-the-shelf LP solvers can then be used to find solutions to the resulting model where the computed solution constitutes the weights in the anonymized graph. As a proof of concept, we choose the shortest paths problem, and experimentally evaluate the proposed techniques using real social network data sets.
Keywords :
"Social network services","Privacy","Tree graphs","Data mining","Facebook","Advertising","Information analysis","Computer science","Linear programming","Shortest path problem"
Publisher :
ieee
Conference_Titel :
Data Engineering (ICDE), 2010 IEEE 26th International Conference on
ISSN :
1063-6382
Print_ISBN :
978-1-4244-5445-7
Electronic_ISBN :
2375-026X
Type :
conf
DOI :
10.1109/ICDE.2010.5447915
Filename :
5447915
Link To Document :
بازگشت