Article ID: | iaor1994411 |
Country: | United Kingdom |
Volume: | 2 |
Issue: | 3 |
Start Page Number: | 225 |
End Page Number: | 236 |
Publication Date: | Jul 1993 |
Journal: | European Journal of Information Systems |
Authors: | OKeefe R.M., Chung Q.B. |
Keywords: | artificial intelligence: decision support |
Model management systems (MMS) have typically focused on model formulation or problem representation, and thus have often been conceived as tools for experienced model builders. In this paper, attention is drawn to the roles of naive users of MMS. Then the authors present a prototype MMS which was designed and developed with this rationale. MMS II, a frame-based prototype MMS, is a tool equipped with facilities for direct retrieval of models, guided retrieval, model modification and integration of models retrieved from different model bases. Discussed here are the original features designed to aid naive users, i.e. non-professional model builders. These include navigational guidance through a model base, if-sought attachments for explanation and suggestions, providing a critique via shadow processing, approximation following failure to retrieve a suitable model, suggestions of models to modify, and model integration via specification of model links.