Default reasoning on the hypermedia

The nature of a hypermedia-based knowledge-based management system (KBMS) admitting default knowledge is examined. The associative reasoning of hypermedia has been utilized to provide both probabilistic and logical interpretations to default rules. The knowledge acquisition process for one such system is challenging, because it must be done under a controlled experiment permitting Bayesian interpretation of default facts and composition of hard rules in logical form. This obviously entails a very large knowledge base. Both the procedural (logic) and empirical (probabilistic) configuration of default situations on the hypermedia help in focusing only a limited subset of the whole knowledge base. However, the success of the KBMS designer will depend on the power of the abstraction tools provided