Back in November, we spoke with Michael Lang Sr. of Revelytix about the collaboration between business and Semantic Web tech-types that should be part of the development of OWL ontologies, if they’re to be of any real value to enterprises. (That story is here.) Over at its knoodl.com site, Revelytix has communities of ontology developers and subject matter experts joining together, matching what the latter set knows about the domain with what the former set knows about creating, managing and analyzing RDF/OWL descriptions.
Now Revelytix has introduced OntVis, a visualization tool to show the semantic contents of OWL DL ontologies for knoodle.com that is another building block for seamless collaboration between ontologists and business subject matter experts. With OntVis, ontologists now “can communicate the precise semantics of any domain to a subject matter expert or business person,” according to a post on the company’s web site. It’s now possible to explore classes and properties in OWL ontologies and the relationships between resources without ever having to see a line of RDF, as the video below explains. Visualizations can be exported to PNG or PDF files, which can be given to SMEs and business people to review and offer comments.
“The problem we are solving for the enterprise (and the semantic web) is the ability to accomplish sophisticated analytics in a federated information environment,” Lang tells The Semantic Web Blog. He goes on to explain that the typical data warehouse or application database of today locates data in the same place as the query engine, and there is a fixed schema. “The relations between concepts are very precise and coded into the DBMS,” he says. “In a federated world, a schema is still required to produce the type of analytics required by the use case; this schema exists as an OWL model. If the OWL file does not represent a domain precisely, then the analytics will not be correct.”
While OWL modelers some day will become experts in the domains being modeled as more businesses take advantage of their expertise, today that is not generally the case. And, if you don’t have subject matter experts participating in ontology makeup, it won’t accomplish the mission. So, in order to get the OWL model of a domain correct, subject matter experts must validate the semantics that are incorporated into the model, Lang says. “These experts cannot read an OWL document directly because it is essentially a foreign language to them,” he says. “They are used to reviewing visualizations of models and commenting on the representation given by the visualization.
OntVis is the first visualization capability that present OWL DL semantics precisely and is readable by subject matter experts and business people.”