Privacy preserving local analysis of digital trace data: A proof-of-concept

TL;DR

PORT is a software platform that allows individuals to locally process their digital trace data, enabling privacy-preserving social data analysis by preventing sensitive data exposure to outside parties.

Contribution

The paper introduces PORT, a novel software tool that implements local data processing of digital traces, enhancing privacy in social data science research.

Findings

PORT successfully processes data locally on user devices.

It shields sensitive data from external access during analysis.

Enables new social science applications with privacy protection.

Abstract

We present PORT, a software platform for local data extraction and analysis of digital trace data. While digital trace data collected by private and public parties hold a huge potential for social-scientific discovery, their most useful parts have been unattainable for academic researchers due to privacy concerns and prohibitive API access. However, the EU General Data Protection Regulation (GDPR) grants all citizens the right to an electronic copy of their personal data. All major data controllers, such as social media platforms, banks, online shops, loyalty card systems and public transportation cards comply with this right by providing their clients with a `Data Download Package' (DDP). Previously, a conceptual workflow was introduced allowing citizens to donate their data to scientific- researchers. In this workflow, citizens' DDPs are processed locally on their machines before they are asked to provide informed consent to share a subset of the processed data with the researchers. In this paper, we present the newly developed software PORT that implements the local processing part of this workflow, protecting privacy by shielding sensitive data from any contact with outside observers -- including the researchers themselves. Thus, PORT enables a host of potential applications of social data science to hitherto unobtainable data.