Ontology Based Construction of Bayesian Network

In ontology engineering tasks such as modeling of domain, ontology reasoning and mapping of concepts between ontology, dealing with uncertainty is vital. Bayesian networks are used to determine the likelihood of occurrences affected by different factors. In this paper the concept of ontology to Bayesian network conversion requires the following tasks: i) determining the relevant nodes, ii) relationships between the factors recognized (connections), iii) for each node with in the Bayesian network calculates the CPT’s. Bayesian networks can capture and enhance the mixed way interdependence between techniques for ontology mapping. We outline the basic idea behind our approach and show some experiments on upper ontology. Then will focus on probabilistic ontology and their relationship with Bayesian Networks. At last it shows how the uncertainty can be handled by Bayesian network notion.