Models of temporal dependencies for a probabilistic knowledge base

Abstract

The article presents models of temporal dependences for constructing probabilistic temporal rules in the Markov Logical Networks. Such rules describe the relations between the states of a control object and taking account the possibility of integrating different approaches of management according to the paradigm of “Enterprise 2.0” knowledge sharing. The proposed models define constraints and conditions for changing the states of a control object, which allows predicting possible variants of its behavior in relation to the current state and providing decision support based on a choice of the most likely variants.