Research

Vision

Our research focuses on the general subjects of data management, data logistics / process support, and data analysis. In particular, classical database topics like data modeling, efficient data access, and concepts for the guaranty of data consistency as well as advanced themes like continuous management and optimization of data quality are in the focus.

In the following, the research foci will be characterized in more detail.

Data Management

Database systems enable an efficient management of structured data, but they have deficits in dealing with special kinds of data, like data streams, electronic documents, and weakly structured data. The Chair investigates well-known and new methods of data management to efficiently apply them to these kinds of data. The project CoBRA DB provides a modular database management system for the implementation of these methods. It allows adapting data management to specific needs and unusual technological environments.

One highlight is the handling of data streams. The objects of interest are special data records with timestamps that represent events generated in large numbers e.g. in sensor networks. Data-Stream Management Systems (DSMS) support the efficient handling of such data. The Chair develops cost models for queries on data streams that can be used for the optimization of data-stream processing. In particular, the project DSAM (Data-Stream Application Manager, i6sdb) strives for the automatic linkage of heterogeneous DSMS in order to exploit the strengths of each individual system. Furthermore, the improvement of data quality in data-stream systems is investigated. A specific application area that also has to do with data streams and event processing is defined by the so-called MMOGs (massively multiplayer online games). Here, extremely many players interact simultaneously in a shared virtual world. The state of that virtual world continuously changes because of the decentralized generation of events. In the context of the project i6engine, the improvement of performance and scalability is investigated with the help of overlay networks.

Database systems today offer only limited support for the preservation of data quality. To cross the borders of individual database systems in guaranteeing high data quality in information systems, new methods and tools are required to support an encompassing Data-Quality Management in an appropriate way. The research project DQ-Step analyses data-quality problems in the context of Concurrent Engineering and develops methods and tools for the sustainable optimization of data quality, while taking legacy systems into account. This intends to identify data-quality problems as early as possible in the data-production process or to avoid them in the first place by installing suitable procedures that support engineers in monitoring and managing their own data quality needs.

Data Logistics and Process Support

Database systems also play a major role in application integration. The kernel of each integration project is the data integration which requires on one hand the semantic mapping and on the other the cross-system synchronization. It provides the prerequisites for a suitable process-oriented integration that optimizes data logistics with reference to the needs of business processes. Data must be exchanged and kept consistent among applications and the databases used by them. To achieve that, processes are defined and executed with computer support ("workflow management", "process management"), in which data must often be transformed to other application-specific formats ("data transfer and conversion"). Here, the semantic integration of data types and instances requires a substantial manual effort. It is a must to search for methods and technologies that minimize this effort. An important constraint is the fact that commercial information systems undergo permanent change. The IT infrastructure of an enterprise must not inhibit this change, but must support organizational learning. In the context of research on evolutionary information systems the Chair deals with the issue of how information systems can be designed to minimize the effort for demand-driven system evolution. Some important principles that contribute to improve the ability of distributed information systems to evolve on demand are among others the deferred design, the separation of concerns, and the loose coupling of components.

The project ProMed ("Process support for adaptive evolutionary information systems in medicine") is set to find out how the communication partners in medical supply networks (hospitals, practicing physicians, laboratories, health insurance companies, and drugstores) can be decoupled in inter-institutional processes with the help of active documents, while at the same time a flexible process support can be enacted.

Data Analysis

The requirements from data-management systems for data analysis (OLAP) are fundamentally different from those from operational database systems (OLTP). So-called data-warehouse systems are established as the technological basis for data analysis. The Chair investigates the specific requirements from the analysis of special data types and how to meet these requirements. Furthermore, an evaluation is done on how the well-known methods of data analysis and data mining can be utilized for the optimization of data quality. A running endeavor in this area is the project Pixtract. The task here is to annotate automatically a large set of images with content-oriented metadata that can be obtained through object-recognition methods.

More Information

• Projects
  Some current and finished internal projects in the last years
• Cooperations
  Some current and finished projects with external partners (cooperations) in the last years
• Fellowships
  Links to different foundations