Welcome to the research group for Data Management and Engineering (dme).

?

Our current research deals with the design and evaluation of algorithms for the integration and synthesis of relational data with a special focus on schema matching and entity matching.

?

Our teaching covers both algorithms and systems for data management and processing. In addition to data integration and data synthesis, this also includes dataset discovery, data profiling, data cleaning, data mining, relational and non-relational database management systems, as well as distributed data processing and the processing of real-time data.

Two of our research papers are accepted at this year's International Conference on Extending Database Technology (EDBT).?Both research projects are?conducted in cooperation with the Hasso Plattner Institute in Potsdam.

?

• PRISMA: A Privacy-Preserving Schema Matcher using Functional Dependencies.?To integrate data sources, correspondences between their schemas must first be identified. Such identification proves to be difficult if the data sources to be compared are encoded differently (e.g. different languages or different encryptions). In this paper, we present PRISMAA, a novel encoding-independent schema matcher that uses functional dependencies within the individual data sources to match their schemas.
  ?
• Icewafl: A Configurable Data Stream Polluter.?The benchmarking of data engineering algorithms requires suitable test data. In many cases, data errors play a major role (e.g., to evaluate the effectiveness of algorithms for detecting such errors or the robustness of data processing algorithms against such errors). In this paper, we present a novel approach to data stream pollution.