Ontology-Based Interestingness in COVID-19 Data

This paper describes an approach to mining interestingness in data by designing domain ontology, COKPME and populating it with anonymized COVID-19 data from private hospital in Karnataka State, India. In general, association rules applied to healthcare data generate a large number of rules. These generated rules may not guarantee interestingness of the generated knowledge. To address this, we propose an ontology-based interestingness measure using the association rule mining algorithm. With the association rule, the implicit relationship between different categories of data attributes is captured. Our approach is to design the domain ontology, populate with data instances and operate association rules for semantic and non-semantic data to discover interesting facts.