In cooperation with the University of Rostock, the DFG project “Interval methods for robust predictive-model based control of solid oxide fuel-cell systems” studies ways of analyzing the dynamic behaviour of solid-oxide fuel cells. Since it is not possible to measure the temperature within a fuel cell stack at all points, the parameters of the differential equations used to describe the stack are inevitably uncertain. To analyze this uncertainty, the project partners from the Faculty of Engineering developed “Vericell”, a new interoperable model and simulation environment. While in the first phase of the project the main interval methods were made available, the implemented middleware permits easy integration of new models and alternative methods of analysis.
Researchers working on the EU project “Knowledge Helper for Medical and Other Information Users” (KHRESMOI) are developing a multilingual and multimodal data system for medics and laypeople to support searches for biomedical information and documents. One focus has been on investigating the influence of different interface components on search behaviour and success. Searches for more specialist information, such as the similarity search for medical radiological images, are also addressed by adapting the user interfaces. The partners of the DFG project “Semantic Cluster analysis at Information Retrieval” (CAIR) develop new methods for document clustering and evaluate them from the user’s perspective. They were thus able, for the first time, to formulate a theoretical model for ideal clustering, creating a solid base for the previous heuristic approaches and at the same time revealing ways of achieving better clustering methods. The DFG project “Supporting collaborative multimedia based learning” investigates mechanisms in the presentation of know­ledge-based information through learning partners. Other projects address stress, decision-making and pathological buying behaviour or tests as ­refresher intervention in preserving skills during complex work activities over lengthy periods of time. As part of the DFG project entitled “Highly interactive information retrieval” (HIIR), eye trackers and system logs are used to monitor users as they search in order to build a stochastic model of their behaviour. In addition to improving the search system, the model helps to guide users during searches. Scientists from the Department of Computer and Cognitive Sciences made a successful contribution within DFG Priority Programme 1409 “Science and the Public” with their research project looking at the determinants of selection and attitude creation while receiving scientific information on the internet.
In collaboration with the Erwin L. Hahn Institute for Magnetic Resonance Imaging, research has also been conducted on addiction to the internet, stress and decision-making.