Knowledge Management for Engineering Design Search and Optimization (EDSO) addresses issues in the process from knowledge acquisition, modelling to knowledge reuse, sharing and publication. We use state of the art technology from Ontologies and the Semantic Web as well as traditional best practices such as rule based reasoning and templates framework, etc. The purpose is to allow access to the whole knowledge lifecycle in the Geodise project.

An Ontology is the conceptualization base for knowledge. Knowledge acquisition has been done on domain experts and relevant documentations to extract this conceptualization in the form of a hierarchy of domain vocabulary and their relationships. This ontology includes a task ontology, optimization ontology, user ontology and application ontology. The task ontology defines key task concepts and their input and output in EDSO to facilitate workflow construction as well as a rule based reasoning workflow advisor. The optimization ontology formally defines the vocabularies in describing optimization methods and their configurations aiming to provide knowledge guidance in optimization selection and configuration. The user ontology focuses on the user profile. Gambit, as a CAD meshing tool, is the target for application ontology which defines commands and parameter names needs for an ontology assisted Gambit script editor. The editor can help engineer editing Gambit journal file by providing context help on commands and parameters.

Geodise uses protégé and OILED as the ontology editors and the ontologies are delivered to our Workflow tool using a generic web/ grid- service enabled ontology service: it is this which we will place into the public domain.

Our workflow construction environment uses the task ontology and allows a user to build workflow by selecting and configuring task components based on the ontology files. The workflow is to be converted to a valid Matlab “.m” file that can be submitted to the MATLAB engine for execution. Other details can be found in the workflow tool description above.

While the Workflow tool provides a mechanism to build a workflow at a conceptual level, conversion and execution need further refined resources. We use the semantic web approach to model and expose various resources in EDSO. Resources, such as “.m” files and problems, can be modelled using RDF triple statements as well as XML files. The RDF is backed with RDF schema (RDFS) as well as a corresponding DAML + OIL ontology (moving to OWL soon) as a complement for further reasoning for resource classification. The ultimate purpose of the semantic web approach in Geodise is to allow engineers to manage their resources easily (by which we mean compute, application codes, and data services), make them searchable and reusable as shared resources in a grid enabled environment.

Geodise Specific Services