Privacy preserving event sequence data visualization using a sankey diagram-like representation

Given the growing rates and richness of data being collected nowadays, it is non-trivial for data owners to determine a single best publishing granularity that presents the most value of the data while preserving its privacy. There have been extensive studies on privacy preserving algorithms in the data mining community, but relatively few have been done to provide a supervised control over the anonymization process. We present the design and evaluation of a visual interface that assists users to employ commonly used data anonymization techniques for making privacy preserving visualizations of the data. We focus on event sequence data due to its vulnerability to privacy concerns. Our visual interface is designed for data owners to examine potential privacy issues, obfuscate information as suggested by the algorithm, and fine-tune the results per their requests. Case studies using multiple datasets under different scenarios demonstrate the effectiveness of our design. These studies show that using visualization as an interface can help identify potential privacy issues, reveal underlying anonymization processes, and allow users to balance between data utility and privacy.