hao-ontology

The Health Analysis Ontology (HAO) captures the data domain and analysis workflow in cartridges that represent domain knowledge, analytics knowledge, and other components of the analysis pipeline. We leverage existing and novel ontolo-gies to formally specify concepts and relationships. In the HAO,this is done by declaratively encoding background and analysis domain knowledge in cartridges and then instantiating them in the knowledge graph. The KG captures all domain knowledge necessary for the use of NHANES and the specification of the population health study. The cartridges allow many different components of data science workflows to be automatically generated at execution time. Example components include the definition of the problem, filtering criteria for the data, selection of the modeling method, execution of the analysis, and presentation of the results.