Online Program Home
My Program

Abstract Details

Activity Number: 475 - Understanding Threats to People, Data, and Privacy
Type: Contributed
Date/Time: Wednesday, July 31, 2019 : 8:30 AM to 10:20 AM
Sponsor: Social Statistics Section
Abstract #306904
Title: Protecting Privacy of Household Panel Data
Author(s): Shaobo Li* and Matthew Schneider and Yan Yu and Sachin Gupta
Companies: University of Kansas and Drexel University and University of Cincinnati and Cornell University
Keywords: privacy; marketing; choice model; synthetic data
Abstract:

We investigate the vulnerability of widely-used household panel data to intruders who intend to re-identify panelist and learn their private information by linking to external databases. We find evidence of high vulnerability and propose a framework to protect household panel data without reducing their commercial value. That is, retailers and manufacturers are able to use the household panel data to perform common analyses such as models of brand choice, but not re-identify individuals by linking to external databases. The proposed protection method is based on a Bayesian hierarchical multinomial logit model that synthesizes the brand choices at various protection levels while preserving household heterogeneity, which is crucial for choice modeling and other marketing analyses. We also compare our proposed protection methods with other common data protection approaches, and show the advantages. We demonstrate the usefulness of our framework using IRI household panel data for multiple product categories.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program