| Research Projects | Promoted | Differential Privacy: New approaches for managing social big data
bidt background

Differential Privacy: New approaches for managing social big data

The research project explored how valid statistical conclusions can be drawn without violating individual privacy. As part of the project, a software environment for implementing differential privacy was developed.

Project description

The Internet of Things (IoT) has given rise to vast databases containing personally identifiable information, forming the basis for new business models: platform operators generate a significant portion of their profits through the analysis of personal data. In addition to the threats posed by uncontrolled data leakage (such as the Cambridge Analytica case in 2018) and data breaches, these big data collections and associated analytical practices spark new challenges for data protection, focusing on individual privacy and the general handling of personal data. A new scientific method that integrates data privacy, confidentiality, and data analytics is “Differential Privacy”: A problem associated with traditional anonymization approaches to personal data is that, while they obscure the identity of the person in the data, they often retain the characteristics and attributes contained within it. In contrast, the Differential Privacy method modifies all personal characteristics and attributes in such a way that individual privacy is preserved while still enabling valid statistical data analysis. This is primarily achieved by injecting noise into statistical analyses using various mathematical techniques.

Our project, “Differential Privacy: A New Approach to Deal with Social Big Data,” provided the first comprehensive evaluation of this new approach and yielded key insights for further development, governance, and implementation. Our findings culminated in policy papers, scientific publications, and policy advisory. Furthermore, a learning platform was developed where users can learn about the concept of Differential Privacy and gain insights into its background. Based on the experience gained from creating the learning platform, the project concluded with a feasibility study to assess whether it is viable to establish a dedicated Differential Privacy platform through which scientific data can be provided to researchers in a privacy-compliant manner. The project approached the topic through four work packages to answer the following questions:

  • What are the opportunities of Differential Privacy from technical and societal perspectives?
  • What are the risks of Differential Privacy for users and society?
  • How can Differential Privacy be practically implemented from a software engineering perspective and economically exploited?

The research team concluded that a “one-size-fits-all” solution for handling Social Big Data is not effective. They identified educating about privacy protection technologies as a core element and were able to formulate new metrics and GDPR requirements on a legal level. The project demonstrated that Differential Privacy holds high potential but faces many implementation hurdles.

The Project was completed on February 28th, 2023.

Contact

Dr. Christoph Egle

Managing Director, bidt

Project team

Prof. Dr. Simon Hegelich

Professor of Political Data Science, Bavarian School of Public Policy | Technical University of Munich

Fabienne Marco

Research Assistant, Chair for Political Data Science | Hochschule für Politik, Technical University of Munich

Andree Thieltges M.A.

Research Assistant, Chair for Political Data Science | Hochschule für Politik, Technical University of Munich

Prof. Dr. Florian Matthes

Chair for Software Engineering for Business Information Systems | Technical University of Munich

Oleksandra Klymenko M.Sc.

Research Associate, Chair for Software Engineering of Business Information Systems | Technichal University of Munich

Gonzalo Munilla Garrido M.Sc.

Research Assistant, Chair for Software Engineering for Business Information Systems | Technical University of Munich

Sascha Nägele M.Sc.

Research Assistant, Chair for Software Engineering for Business Information Systems | Technical University of Munich

Prof. Dr. Uwe Baumgarten

Chair for Computer Science F13 Department & Data Protection Officer | Technical University of Munich

Dmitry Prokhorenkov

Doctoral Candidate and Research Assistant, Chair for Computer Science | Technical University of Munich

Resources