Towards adaptive open-world modeling

In this technical report, extensions to the Object-Oriented World Model (OOWM) are proposed allowing for adaptive open-world modelling for artificial cognitive systems. In cognitive systems, a world model can serve as a central component for integrating, storing and disseminating information from an observed environment. Thus, a world model depicts an abstract, simplified representation of an observed real-world domain. For allowing high-level information processing on a semantic layer, representations of real-world entities, which are observed by sensors, can be semantically enriched by domain models. In general, a domain model contains only a fixed number of a priori defined concepts from a closed world. However, in many real-life applications, the considered environment is not closed. For coping with changing environments, a cognitive system must be equipped with an adaptive world model able to adjust to an observed open environment. Thus, this technical report discusses how the OOWM can be extended to facilitate semantically grounded open-world modeling.