A Bayesian network approach to handle uncertainty in Web Ontology Language. (17th September 2021)

• Record Type: Journal Article
• Authors: Malik, Sonika; Jain, Sarika
• Is Part Of: International journal of reasoning-based intelligent systems. Volume 13:Number 4(2021)
• Page Start: 243
• Page End: 250
• Publication Date: 2021-09-17
• Subjects: ontology -- uncertainty -- Bayesian network -- unit of knowledge; Artificial intelligence -- Periodicals; Reasoning -- Periodicals; 006.3
• Languages: English
• ISSNs: 1755-0556

Abstract

In information management systems ontology play a vital role. An ontological application should include a mechanism for handling uncertainty. Ontological innovations are to become the web's future in the coming period, but there are still some features such as exceptions, vulnerability and default values missing. Ontological languages such as OWL and RDF are by necessity distinct in nature, so ambiguous details cannot be treated. In this research, article uncertainty is handled in the ontology by Bayesian network. A probabilistic model of uncertainty available in the knowledge base is the Bayesian network. The Bayesian network is a conceptual model that is ideal for the representation and analysis of ambiguity and information found in data. The probability of uncertainty can be used for many real-life scenarios in the knowledge-base. We also introduced defaults and exceptions along with uncertainties to enhance performance and improve the OWL features. The source code is then translated to a jar-file with maven and it can be used in Protégé itself.