Semi-automated Ontology Creation and Upgradation for Rail-Road Incidents: A Case of a Steel Plant in India

Abstract

Most of the earlier studies for ontology creation and upgradation are carried out using structured data. Very few of them focused on unstructured texts for this purpose, which are restricted due to huge human involvement. In order to address this issue, in this study, we have proposed a methodology for ontology creation and upgradation using both unstructured (text) and structured data of rail-road incidents obtained from a steel plant. Features are predicted from the texts using machine learning (ML) techniques and added to structured data, which is finally used for ontology upgradation. This is performed in a semi-automated way, thereby substantially reducing level of human efforts. Ontology is created using Web Ontology Language (OWL) language in Protege software. Two ML techniques, Random Forest and Support Vector Machine are used to see how accurately the logical relationships are established between classes/sub-classes used in this ontology. Hence, the study proves the effectiveness of the methodology for ontology creation and upgradation.