Ontology-Based Reasoning with Uncertain Context in a Smart Home: A Decision Network Approach

Smart homes present complex domains that suit ontological modelling. To be efficient, smart home systems must incorporate domain uncertainty into decision making. But ontologies presently cannot represent and reason under uncertainty. It is thus challenging, developing ontology-based decision-making models that can integrate domain uncertainty. In this paper, we seek a decision network approach as a probabilistic extension to OWL ontology using only subset of domain concepts relevant for probabilistic modelling. We introduce an algorithm, OWLMB, which can generate this relevant set of concepts in ontology. Also, we propose representations for decision and utility nodes for valid decision situation in ontology. Validation of this approach in a smart home scenario shows our approach can perform well in a real application domain.