IDFinder: Data Visualization for Checking Re-identifiability in Microdata
Hyunmo Kang, Sam Hawala, Laura Zayatz
KEY WORDS:
ABSTRACT
This paper introduces a novel user interface, IDFinder, which is specifically designed to facilitate the disclosure avoidance process on microdata files. IDFinder is designed based on the well-known visual seeking mantra, "Overview first, Zoom and filter, and Details on demand." Direct data manipulation and dynamic query techniques implemented in IDFinder provide users rapid, incremental and reversible operations which are critical for disclosure avoidance tasks. Multiple tightly coupled data viewers are used to represent the different data hierarchies in microdata. In addition, time series viewers which are also tightly coupled with other data viewers, visualize the change of attribute values over time and enable users to observe the attribute values in each data hierarchy at the specified time. The usability study with a small group of disclosure avoidance researchers led to the refined designs of IDFinder, and it also revealed benefits, scalability issues, and applicability to other tasks.
CITATION:
