A data visualization tool for finding optimal privacy- utility trade-off in data analysis
Many enterprises make use of data analysis. However, in order to fulfill the demands of different privacy levels and information utility in different scenarios, data owners and data consumers must share data in a privacy-preserving way. Several anonymization techniques have been developed and used for privacy protecting but few in the aim of data visualization. This study proposes a visualization tool for anonymized data that allows users to change privacy and utility levels in different usage scenarios. The tool includes standard data analysis charts and diagrams, and by altering the level of privacy, users may discover and reduce privacy breaches in the visualization during the iterative process. This paper uses two case studies with real-world datasets to demonstrate how this approach helps discover and resolve potential privacy issues while balancing overall visualization readability and utility. Think-aloud is applied to evaluate the utility of the prototype which can be helpful for upcoming work on anonymous visualization.