Geolocation with respect to personal privacy for the Allergy Diary app - a MASK study

Area of Science:

• Health Informatics
• Data Privacy
• Mobile Health

Background:

• User localization data is crucial for the MASK (Mobile Airways Sentinel networK) app.
• Geolocation data is classified as personal data, necessitating robust privacy measures.
• Effective data anonymization is essential for maintaining user privacy in mobile health applications.

Purpose of the Study:

• To evaluate the effectiveness of different anonymization techniques for user geolocation data within the MASK app.
• To compare the privacy protection levels offered by noise addition (randomization) versus k-anonymity (generalization).
• To identify an optimal privacy preservation method for the MASK database.

Main Methods:

• Analysis of geolocation data from MASK app users between December 2015 and November 2017.
• Implementation and comparison of two anonymization methods: noise addition and k-anonymity.
• Assessment of data accuracy and privacy protection capabilities of each method.

Main Results:

• Over 93,000 user-days of data were collected, with 58.5% successfully geolocalized.
• K-anonymity demonstrated superior accuracy in protecting user privacy compared to noise addition.
• The study identified k-anonymity as a more effective method for safeguarding sensitive geolocation information.

Conclusions:

• K-anonymity is a suitable and effective method for anonymizing MASK app data.
• The findings suggest that k-anonymity can be applied to other databases requiring similar privacy protections.
• Implementing k-anonymity enhances the privacy of user geolocation data in mobile health research.