Reasoning with Generalized Cases

Similarity-based retrieval of stored objects is one of the main building blocks of a case-based reasoning approach (CBR). Over the last three decades several sophisticated retrieval methods have been developed which are more suitable for many real-world scenarios than the traditional database search. However, the CBR approach and in particular the published retrieval methods do not support sufficiently special “complex” objects called generalized cases. My Ph.D. thesis [6] clarifies the concept of generalized cases and describes application scenarios. It discusses representation approaches and then focusses on novel methods for similarity assessment and efficient retrieval.