Case-based reasoning (CBR) is an established problem solving paradigm from Artificial Intelligence (AI). It is built upon a rule of thumb suggesting that similar problems tend to have similar solutions. More specifically, the idea of CBR is to exploit the experience from similar problems in the past and to adapt their successful solutions to the current situation. Thus CBR implements experience-based problems solving.

The core of every case-based problem solver is the case base, which is a collection of memorized chunks of experience, called cases. The case base is usually stored in a data bases and constitutes the core knowledge of the problem solver. New problems are solved by retrieving cases from the case base which are similar to the current problem. The experience stored in such similar cases is then reused, for example, solution pieces are adapted towards the new problem and possibly combined. CBR research has developed a considerable set of methods to realize similarity-based retrieval and adaptation of cases. CBR systems are also adaptive systems as they continuously update their case base: new problem solving experience is retained and outdated experience is removed.

Our competences

For further information about CBR, read the following survey article:

Bergmann et al. (2009). Case-Based Reasoning - Introduction and Recent Developments. Künstliche Intelligenz, 1/2009:5--11.

Our competences

We are one of the leading groups on CBR world wide. The head of the research group has more than 20 years of experience in CBR research and development. He published well recognized work on the foundations of CBR, including case representation, similarity assessment, compositional case adaptation, efficient similarity-based retrieval, and learning.

Our current work on CBR focuses on reasoning with experience on the Web, thus making the huge amount of experience in social media, forums, and Q&A web pages accessible to CBR technology. To this end, we develop new methods for the extraction of structural cases from the web. A second area of current research is process-oriented CBR. We address case representation, similarity assessment, and adaptation of experience represented in workflows. An important challenge is to find similarity measures for workflow retrieval and to enable efficient workflow retrieval from large case bases. A third topic of continuous interest is research on novel solution adaptation methods, including hierarchical adaptation, e.g. for case-based planning, compositional adaptation, and a frameworks for learning adaptation knowledge.

  • research on all areas of CBR, including case representation, similarity assessment, adaptation, and learning
  • reasoning with experience on the Web
  • CBR support for workflow systems
  • development of generic CBR tools
  • development of commercial CBR applications