Semi-automatic generation of ontologies for knowledge-intensive CBR

This work examines how automatic generation of general domain knowledge from text can contribute to the knowledge intensive Case-Based Reasoning (CBR) system, Creek. This will be done by combining Creek with the ontology extraction tool CORPORUM to improve Creek’s ability to retrieve text. The ontology extraction tool is part of an architectural framework that is described. The framework describes a knowledge acquisition cycle, which is base on natural language textual documents. The framework seeks to use all the available information of the documents in addition to dictionaries to improve knowledge acquisition process. We continue by implementing parts of the architectural framework in Java.