Organizations often provide workers with knowledge management systems to help them obtain knowledge they need. A significant constraint on the effectiveness of such systems is that they assume workers know what knowledge they need (they know what they don’t know) when, in fact, they often do not know what knowledge they need (they don’t know what they don’t know). A way to overcome this problem is to use visual ontologies to help users learn relevant concepts and relationships in the knowledge domain, enabling them to search the knowledge base in a more educated manner. However, no guidelines exist for designing such ontologies. To fill this gap, we draw on theories of philosophical ontology and cognition to propose guidelines for designing visual ontologies for knowledge identification.
We conducted three experiments to compare the effectiveness of guided ontologies, visual ontologies that followed our guidelines, to unguided ontologies, visual ontologies that violated our guidelines. We found that subjects performed considerably better with the guided ontologies, and that subjects could perceive the benefits of using guided ontologies, at least in some circumstances. On the basis of these results, we conclude that the way visual ontologies are presented makes a difference in knowledge identification and that theories of philosophical ontology and cognition can guide the construction of more effective visual representations. Furthermore, we propose that the principles we used to create the guided visual ontologies can be generalized for other cases where visual models are used to inform users about application domains.