TOPKAT - The Open Practical Knowledge Acquisition Toolkit

Click on this diagram to see a larger version of it.

TOPKAT (The Open Practical Knowledge Acquisition Toolkit) is a hypertext and diagram-based toolkit, developed at AIAI, which integrates knowledge elicitation techniques with the CommonKADS approach to knowledge modelling. It is implemented in HARDY, AIAI's tool for producing hypertext and diagrams in user-defined formats, and in CLIPS. TOPKAT supports the following knowledge elicitation techniques:

Hypertext-based transcript analysis, including a partial natural language analysis which helps in the identification of concepts and properties;
Card sorting, including assistance in classifying sorts as different types of properties;
Repertory grids, including support for transforming numbers into meaning ful names;
Laddered grids, including a facility for creating a CommonKADS domain model directly from a laddered grid.

TOPKAT supports the following CommonKADS models:

Task modelling (partial)
Communication modelling (partial)
Expertise modelling:
- Domain dictionaries
- Domain models
- Model level
- Inference structures
- Library of inference structures
- Task level
Design modelling (basic)

TOPKAT also supports PostScript output of models and diagrams. For more detail, see this file, or read this paper, which was presented at the BCS SGES Expert Systems 94 conference. The current status of TOPKAT is that it is a prototype used in previos projects. No further development is planned.