Metadata-Based Privacy Assessment for Mobile mHealth

Area of Science:

• Health Informatics
• Mobile Health
• Data Privacy

Background:

• Mobile health (mHealth) applications are increasingly collecting sensitive personal and physiological data.
• Current privacy compliance relies on self-declared information, which is unreliable.
• Existing app stores and privacy labeling systems lack robust data-driven privacy assessments.

Purpose of the Study:

• To introduce App-PI, a data-driven ecosystem for managing mobile application privacy.
• To automate the collection, analysis, and visualization of privacy-related metadata from mHealth applications.
• To provide end users with manageable privacy tools and researchers with reliable app metadata.

Main Methods:

• Integration of heterogeneous data sources into a unified repository (App-PIMD).
• Automated collection and analysis of privacy metadata from mobile applications.
• Focus on data flow design for quality assessment of privacy impact data.

Main Results:

• Demonstrated the functionality of App-PI using a popular mHealth application.
• Showcased the importance of data flow design for generating usable privacy insights.
• Enabled empirical assessment of privacy risks associated with mHealth apps.

Conclusions:

• App-PI offers a structured and reliable approach to assessing mHealth app privacy.
• Effective data flow design is crucial for delivering actionable privacy information to users.
• The ecosystem empowers users and researchers with data-driven privacy management tools.