DPVisCreator: Incorporating Pattern Constraints to Privacy-preserving Visualizations via Differential Privacy

Jiehui Zhou, Xumeng Wang, Jason Kamkwai Wong, Huanliang Wang, Zhongwei Wang, Xiaoyu Yang, Xiaoran Yan, Haozhe Feng, Huamin Qu, Haochao Ying, Wei Chen

View presentation:2022-10-20T16:45:00ZGMT-0600Change your timezone on the schedule page
2022-10-20T16:45:00Z
Exemplar figure, described by caption below
The overview of our visual analytics approach for generating privacy-preserving visualizations. Users can select patterns of interest from visualizations of tabular data and specify importance weights. The PriVis model considers these pattern constraints and privacy budgets to generate privacy-preserving schemes. Users can evaluate the results based on the visualization charts and utility metrics and return for iterative adjustments.

Prerecorded Talk

The live footage of the talk, including the Q&A, can be viewed on the session page, Digital Humanities, e-Commerce, and Engineering.

Fast forward
Abstract

Data privacy is an essential issue in publishing data visualizations. However, it is challenging to represent multiple data patterns in privacy-preserving visualizations. The prior approaches target specific chart types or perform an anonymization model uniformly without considering the importance of data patterns in visualizations. In this paper, we propose a visual analytics approach that facilitates data custodians to generate multiple private charts while maintaining user-preferred patterns. To this end, we introduce pattern constraints to model users' preferences over data patterns in the dataset and incorporate them into the proposed Bayesian network-based Differential Privacy (DP) model PriVis. A prototype system, DPVisCreator, is developed to assist data custodians in implementing our approach. The effectiveness of our approach is demonstrated with quantitative evaluation of pattern utility under the different levels of privacy protection, case studies, and semi-structured expert interviews.