Events & Distinguished Lectures
To achieve our goal of understanding and controlling privacy as comprehensively as possible, we established three task forces as a mechanism to foster and ensure cross-cutting collaboration.
These task forces are a mechanism to support project partners to coordinate their research, and work together towards the broad vision, beyond the technical cooperations between individual projects.
The task forces focus on general research themes spanning across projects: data models, privacy notions, and usability.
These themes are orthogonal to the two research project groups, reaching across these areas.
The task forces structure and coordinate research on their overarching themes, fostering exchange and collaboration between the participating research projects.
Task Force 1: Data Models
Coordinator: Gerhard Weikum
Data models, i.e., computer-processable representations of privacy-relevant information within users’ modern digital habitats, are an essential basic ingredient for works throughout much of the collaborative research center.
They form the basic data structures in information extraction and analysis; predictive models build on them to model future information propagation; privacy concepts may refer to them as building blocks in more complex properties; programs in the users’ environment manipulate user data, and program analysis aims to keep track of that manipulation; privacy-aware design naturally builds on concepts pertaining to the information being processed.
The first goal of this task force is to identify, in depth, the commonalities and differences between the data models suitable for these different purposes.
For example, data items in program analysis may be attributes in an app’s input or output, in information extraction may be semantic representations of the content of user utterances in online forms, and in predictive models may pertain to hypothetical pieces of knowledge an adversary may be able to gain.
We need to derive a consolidated view of these distinctions, forming the basis for the second goal which is to cross-coordinate the data models, across the collaborative research center, to the extent possible.
The overall outcome of the task force will be a comprehensive framework for data models in the CRC’s methods and tools, providing the basis for their interaction, between different kinds of privacy analysis, and between privacy analysis and enforcement.
Task Force 2: Privacy Notions
Coordinator: Matteo Maffei
For many levels in the proposed technology, there is an inherent need for rigorous notions for defining user privacy.
For example, reasoning about the privacy effects of users in their daily online interactions calls for a privacy notion that can cope with the increasingly dynamic dissemination of unstructured, heterogeneous user content.
Moreover, it has to live up to the open nature of the Internet, which encompasses
ubiquitously available background knowledge and typically only partial access to relevant information.
A different example: modern privacy-enhancing technologies strive to offer rigorous privacy guarantees to end users by suitable incorporation of cryptographic primitives that ensure that sensitive data stays with the user, or by appropriate sanitization that allows to derive privacy-preserving statistics.
While the latter guarantees and thus the corresponding notions of privacy are typically more stringent because they target a closed setting with full control over relevant data, the former setting will have to strive for weaker notions that may include heuristic behavior or weaker adversarial models.
This task force will investigate the interplay of, and cross-coordinate between, the different privacy notions that will arise in the CRC 1223, and which corresponding computer-processable privacy property representations (languages) will be most appropriate.
It in particular aims to understand the commonalities and synergistic effects between the different notions, where and how they are different, and which hybrid notions will
be most suitable to address cross-cutting scenarios.
We will link to Task Force 1 (“Data Models”), as the privacy notions investigated here will operate upon the data models.
Task Force 3: Usability
Coordinator: Antonio Krüger
A major goal of this CRC is to help regular users to understand, and improve, their privacy.
As such, humans and usable user interfaces form part of the considerations in many of the CRC research projects.
Humans need to understand the outcome of a privacy threat analysis, and they need to understand the enforcement options at their disposal.
Ensuring the quality of these interactions requires the design of good user interfaces, and sometimes requires human subject usability testing.
This task force aims to provide a methodological framework for all projects that rely on user interface components, so that state-of-the-art usability research is uniformly taken into consideration.
The task force will collect and distribute the knowledge of how to design usable user interfaces and how to evaluate them, thus helping to establish a standard for user interfaces designed in the CRC.
To that end, the task force will also incorporate researchers of the interdisciplinary research group Saarbrücken HCI1, one of the major HCI groups in Europe.
This will bring additional expertise into the CRC, external to the CRC but local on
the same campus.
The task force will inform user interface design and tests throughout the CRC.
It will coordinate state-of-the-art usability design and user evaluation standards for privacy research.
Where appropriate, standard methods of usability research will be extended and adopted to fit the particular needs of the CRC.
The task force will maintain a repository of suitable usability design and testing tools.