Data Spaces: Combining Goal-Driven and Data-Driven Approaches in Community Decision and Negotiation Support

In the last decade, social network analytics and related data analysis methodologies have helped big players gain enormous influence on the web, largely due to clever centralistic data collection in major data lakes. In the form of recommender systems, this can also be seen as world-scale group decision support. In our research, we have been more interested in how these kinds of technologies can spill over to smaller-scale communities of interest in the long tail of the internet. Examples include learning communities and open source software development communities of individuals, but also questions of controlled data and knowledge sharing among small and medium enterprises or medical institutions. Especially in the latter cases, we often face strongly conflicting goals that need to be negotiated to mutually acceptable solutions, quite along the original GDSS and NSS visions of Mel Shakun and colleagues. One example is medical research support on rare diseases which raises the need for data sharing across multiple health organizations (not necessarily being fond of each other) in a fully transparent, fraud-resistant research process while preserving best-possible privacy of patient data. We end with a summary of the Industrial Data Space initiative recently proposed by Fraunhofer which aims at architectures, rules and tools for data sovereignty in cross-organizational data management and analytics.