Bayesian Network for Uncertainty Representation in Semantic Web: A Survey

Abstract

Bayesian network is a probabilistic model to represent uncertainty available in knowledge base and using it tremendous works have been done to prove its relevance in uncertainty representation and reasoning using Bayesian inference. Probability can be used to represent uncertainty like prediction information, situational awareness, data and knowledge fusion etc in knowledge base to implement various real life situations. Various approaches based on description logic, object oriented, entity relational, and first order logic have been tried to represent uncertainty successfully. One of them is Multi-Entity Bayesian Network (MEBN) logic to represent probabilistic information and performing knowledge fusion in ontology, which is realized using PR-OWL (Probabilistic Web Ontology Language). This paper aims at giving an overall view, the work carried out so far to represent uncertainty with the help of Bayesian Network in semantic web and a list of works done using MEBN/PR-OWL for knowledge fusion or the representation of uncertainty in semantic web.