Probabilistic Knowledge Processingand Remaining Uncertainty

Information is indispensable in preparing economic decisionspurposefully. In this paper knowledge is representedby a probability distribution. Knowledge acquisitionis realized by the principle of maximum entropywhich guarantees an unsophisticated expansion ofknowledge and provides an unbiased basis for decisionprocesses. Under a mathematical point of view entropymeasures remaining uncertainty and reduction of entropyis knowledge increase (first order uncertainty). Ingeneral, an actual epistemic state might vary due to furtherinformation. Thus, it is necessary to have an instrumentwhich reflects also the potential of distribution’spossible variation. Therefore, we present the conceptof so called second order uncertainty, which allows notonly a ’look’ on an epistemic state, as the entropy does,but also ’looks’ into an epistemic state. First the argumentswill be complemented by a short example andthen they will be illustrated by a real world model in thefield of business to business transactions. All calculationswill be supported by the expert system SPIRIT....