• Michael DeBellis

CODO: An Ontology for Collection and Analysis of Covid-19 Data

Updated: Aug 8

Most of my work with technologies such as OWL in the past has focused on developing models. At the beginning of the year I had an interesting project with a client where I helped them design and implement a large knowledge graph for their internal use. Knowledge graph technology is essentially the same standards as the Semantic Web. However, the term Knowledge Graph is the one that has been adopted by the US IT industry. Also, knowledge graphs focus more on scaling up to very large ontologies and as a result make some compromises on the reasoning capabilities that are available with OWL and SWRL when working on small to medium ontologies in tools like Protégé. In order to scale up to tens of thousands or even millions of triples a special type of database known as a triplestore is required. I've been working primarily with the Allegro Graph product from Franz Inc.


Allegro is incredibly fast and also features many excellent additional capabilities to manage a large knowledge graph. It also features a browser called Gruff that is designed to visualize very large knowledge graphs and also supports a full implementation of the latest SPARQL specification.


I have been developing a knowledge graph with Biswanath Dutta a professor at the Indian Statistical Institute. Professor Dutta and I have taken information from the Indian government on the spread of the Covid-19 pandemic and developed a knowledge graph called CODO to help provide additional semantic knowledge about the data. In addition to OWL we have utilized SPARQL to transform text strings in our input data into appropriate objects and property values in the knowledge graph. We develop and maintain the ontology model in Protégé and WebProtégé and we upload and transform our data using Allegro Graph, Gruff, and SPARQL. Our first paper on this work was accepted to the Knowledge Engineering and Ontology Development (KEOD) 2020 conference: CODO: An Ontology for Collection and Analysis of Covid-19 Data


The ontology can be downloaded from the BioPortal site: CODO Ontology at BioPortal

©2019 by Michael DeBellis. Proudly created with Wix.com