Data Mining Workflow Ontology

Data Mining Workflow Ontology

The DMWF ontology is used to model data mining operators, goals, tasks, workflows, and objects like data tables and prediction models. It is used by the AI planner within the Intelligent Discovery Assistant. For each operator, the DMWF ontology contains preconditions and effects that specify on what data an operator can be applied and how the data will change as an effect of this application. The DMWF can be conveniently edited with eProPlan which adds views to the Protégé ontology editor for editing conditions and effects, for connecting it to RapidAnalytics to visualize data tables stored therein, and to apply operators to test conditions and effects. It can even be used to generate workflow plans directly inside Protégé.

The current online version of the ontology is available here:

It imports various ontologies that contain the actual description of the individual operators. Each component (e.g., a Web service or a RapidMiner Extension) developed within e-LICO defines its own ontology file which is imported into the main ontology.