Description:
Goal of this paper is to develop a meta-model, which provides the basis for developing highly scalable artificial intelligence systems that should be able to make autonomously decisions based on different dynamic and specific influences. An artificial neural network builds the entry point for developing a multi-layered human readable model that serves as knowledge base and can be used for further investigations in deductive and inductive reasoning. A graph-theoretical consideration gives a detailed view into the model structure. In addition to it the model is introduced using the example of large software development projects. The integration of Constraints and Deductive Reasoning Element Pruning are illustrated, which are required for executing deductive reasoning efficiently.