Reachability and Dependency Calculi: Reasoning in Network Algebras

In the field of artificial intelligence, it is important to draw conclusions even if there is no complete information. The mathematical approach of qualitative calculi has so far mainly been used for reasoning about space or time. The paper and talk adapts the quantitative techniques for describing universes which are structured as directed graphs. Areas of application are networks, for example computer networks, travel networks, cash flow between international enterprises and political interactions of states. The paper is mainly concerned with the complexity properties of those network calculi. For two classes of calculi called reachability calculi and dependency calculi, the author determines the complexity of determining consistency or finding a model for various versions of the calculi.