Name of the participant: Stefan Arnold
Description of the IT-research project:
Description of the IT-research project: Collaborative Enterprise Networks (CEN) are defined as a form of coordinated collaboration between several autonomous and formally completely independent companies. One type of collaboration consists of the joint creation of a predictive model for which the own amount of data alone is insufficient. When creating the model, it should be noted that the data used for this purpose may contain sensitive and personal information.
Federated Learning (FL) was created to avoid the exchange of sensitive and personal data. Instead of data, Federated Learning exchanges minimal updates needed to improve a predictive model. As a result, Federated Learning offers significant advantages in terms of data privacy and data sovereignty.
The foundations of Federated Learning were originally developed for use on mobile devices. The aim is to evaluate the use of Federated Learning in corporate networks. To evaluate Federated Learning, a prototype for the classification of tax declarative documents is to be developed and subsequently analysed with regard to acceptance and feasibility.
Software Campus partners: FAU, DATEV
Implementation period: April 2021 – März 2023