Abstract:Enabling accurate analysis of social network data while preserving differential privacy has been challenging since graph features such as clustering coefficient or modularity often have high sensitivity, which is different from traditional aggregate functions (e.g., count and sum) on tabular data. In this paper, we treat a graph statistics as a function f and develop a divide and conquer approach to enforce differential privacy. The basic procedure of this approach is to first decompose the target computation f into several less complex unit computations f1, …, fm connected by basic mathematical operations (e.g., addition, subtraction, multiplication, division), then perturb the output of each fi with Laplace noise derived from its own sensitivity value and the distributed privacy threshold ε i , and finally combine those perturbed fi as the perturbed output of computation f. We examine how various operations affect the accuracy of complex computations. When unit computations have large global sensitivity values, we enforce the differential privacy by calibrating noise based on the smooth sensitivity, rather than the global sensitivity. By doing this, we achieve the strict differential privacy guarantee with smaller magnitude noise. We illustrate our approach by using clustering coefficient, which is a popular statistics used in social network analysis. Empirical evaluations on five real social networks and various synthetic graphs generated from three random graph models show the developed divide and conquer approach outperforms the direct approach.

The estimation of bias and variance in clustering coefficient streaming algorithms

Evolutionary Clustering of Streaming Trajectories

Density-based hierarchical clustering for streaming data

Approximate Clustering On Distributed Data Streams

An Improved Network-Based Recommendation Model Via Inhibiting Algorithm Bias

Distributed Censored Regression over Networks

On Learning Cluster Coefficient of Private Networks

Correlation Estimation for Distributed Wireless Video Communication.

An efficient algorithm for approximate biased quantile computation in data streams.

Summarization and Matching of Density-Based Clusters in Streaming Environments

Towards Metric DBSCAN: Exact, Approximate, and Streaming Algorithms

Variational Bayesian Inference for Robust Streaming Tensor Factorization and Completion

A New Parallel Adaptive Clustering and its Application to Streaming Data

Communication-Efficient and Exact Clustering Distributed Streaming Data

MOStream: A Modular and Self-Optimizing Data Stream Clustering Algorithm

When is it Biased? Assessing the Representativeness of Twitter's Streaming API

A Very Efficient Scheme for Estimating Entropy of Data Streams Using Compressed Counting

Asymptotic distributions of the average clustering coefficient and its variant

Streaming Euclidean $k$-median and $k$-means with $o(\log n)$ Space

Understanding video streaming algorithms in the wild

High-performance Broadcasting Algorithms on Cluster