Document analysis and text mining techniques are used to pre-process documents in information retrieval systems, to extract concepts in ontology construction processes, and to discover and classify knowledge along several dimensions. In most cases it is not obvious how the techniques should be configured and combined, and it is a time-consuming process to set up and test various combinations of techniques. In this paper, we present a workbench that makes it easy to plug in new document analysis and text mining techniques and experiment with different constellations of techniques. We explain the architecture of the workbench and show how the workbench has been used to extract ontological concepts and relationships for a document collection published by the Norwegian Center for Medical Informatics.
展开▼
