Towards support in building qualitative knowledge models

Qualitative Reasoning (QR) formalisms provide ontological primitives for capturing conceptual knowledge. Recently QR-based diagrammatic tools are being developed to support learners in creating concept maps as means to acquire such knowledge. However, QR formalisms are complex which hampers their usability. While other approaches address this by simplifying the formalism they use, we seek the solution in providing a set of agents that can support the learner. Based on a previously reported study on using QR modelling tools, we have developed a multi-agent approach to support the QR modelling process. The agents provide different kinds of help, such as general information on the formalisms and tailored feedback addressing the individual needs of a learner. Agents thus have scope, provide context-sensitive help, and are personified according to the type of support they provide. An evaluation study shows that the help-system is well accepted by learners.