Legal and Quality-Centered Requirement and Design Patterns for Context-Aware Applications

• Applicants: Privatdozentin Dr. Silke Jandt; Professor Dr. Jan Marco Leimeister; Professor Dr. Alexander Roßnagel; Professor Dr. Matthias Söllner
• Subject Area: Software Engineering and Programming Languages ; Public Law
• Term: from 2017 to 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 348084924

Context-aware applications are becoming increasingly popular. Especially the combination with characteristics such as machine learning, anthropomorphism, and multimodality enables completely new types of artificially intelligent (AI) systems. Smart personal assistants (SPAs) from providers such as Amazon and Google are successfully making these features available to the broad consumer market. However, the manifold potentials of these systems are also confronted with risks. Reports of discriminatory or incomprehensible system behavior are increasing and contribute to a negative image of AI in society. The DFG-funded, interdisciplinary research project AnEkA addresses this problem. On the basis of legal and quality goals, requirements are derived for legally compatible and high-quality SPAs, conflicting goals are resolved, and design alternatives are identified and evaluated. As a result, requirement and design patterns for the design of legally compatible and high-quality SPAs and AI systems are developed. In order to further increase the economic, technical and social benefits of the patterns, an additional interdisciplinary evaluation with requirements analysts, system developers and users as part of project continuation is of particular importance.

• DFG Programme: Research Grants