Janus* “Automatic” Ontology Builder |
![]() |
During my contributions to the EDIFRANCE and UN/CEFACT harmonization groups I observed that the most part of time was spent in choosing the right name and the right structure of business messages core components, in the aim to be as “general” as possible. Thus compliant with the higher number of existing definitions provided by others groups or standards. The result was always the same, names and structures were always different, and this anytime we developed a new component. Differences were not so relevant but anyway applications implementing defined libraries were not able to interchange between them on the fly, and this sometimes for few details. For this reason I started to study semantics solutions in order to provide a tool able to facilitate information exchanges between applications. With my great surprise rare are existing solutions able to integrate heterogeneous information from XML files, which is the most widely used language format for applications message definition. For this reason I started to develop a tool capable to acquire knowledge from XML schema files and produce semantic information, that nowadays we can call ontologies. A final and stable release of this tool is not ready yet, but firsts result start to become of interest. Here a picture of the tool, presenting a graph automatically generated from a subset of B2B standards for the address concept.
I forgot to say that Janus has been presented at the Developer Track Session of the
17th International World Wide Web Conference (WWW2008). Beijing, China, April 21 - 25, 2008. Documents about Janus *As
often, computer scientists are not so genial to get a name to their own software. ...I belong to the second category. Janus originally was the “Roman god of gates and doors, beginnings and endings”, now is a tool for building ontologies. Probably Janus, the real one, and romans should not be happy to know that, but anyway it is too late now...).
|