Information Filters

The information filter project focuses on improving methods and algorithms in large-scale information filters, such as publish/subsribe systems and message brokers.

Most recently, our research has been directed towards:

• Batching for information filters: Handling many data items or opeations in bulk has been successfully applied on database problems, such as bulkloading of values. For information filters, we evaluate the benefits of handling messages in batches.
• High update rates information filters: Existing information filters have been tuned towards handling large profile numbers and high messages rates by using elaborate indexes. As a results, such systems are not able to handle high or varying update rates gracefully. Thus, we look into methods to handle such update loads better.

Note: we are always looking for motivated students that want to write a Semesterarbeit or MSc thesis in an exciting research project. You should have sound programming skills and team-oriented working style. If your are interested, please have a look at the thesis topics we offer. Do not hesitate to contact us.

People: Donald Kossmann, Jens Dittrich, Peter Fischer