GeoWorlds Project, Distributed Scalable Systems Division
University of Southern California Information Sciences Institute

The Semantic Description Tool allows system engineers to prepare semantic-level  metadata about software components that include descriptions of the functionality, input/output data semantics, and other core properties of the components. By using the tool, ontologies of functionality and data semantics can be defined, and individual metadata schemas can be edited.

The following picture shows a screen shot of the metadata editor. The upper part of the window is for editing service components (analytic, visual, information source, user input, and data conversion services).  The lower part is for editing content and structure semantics of data. In both cases, the left side displays the ontology hierarchy and the right side provides forms to edit schema or instance property values.

1     Define functional and data ontologies (abstract components)

 To register a service component to the GeoWorlds system and make it usable by the service composition tool, the user needs to select or define appropriate functional and data ontologies where the service component and its input/output data descriptions can be classified.

1.1      Add a new functional ontology

To add a new functional ontology, the user needs to decide the type of the service among the GeoWorlds' five top-level service categories: analytic, visual, information source, user input, and data converter services. After selecting a corresponding tab, the user needs to decide the position of the new ontology within the taxonomy hierarchy. A new ontology can be created and located under the selected position by pressing the New Schema ( ) button. Then, the form to edit the schema of the new ontology will be appeared at the right-side panel. Using the schema editor, the user can specify the URI (Uniform Resource Identifier), add a textual description, add more subsumption relations with other ontologies, and define input/output data semantic descriptions.

1.2      Add more subsumption relations

The schema editor allows users to define more than one subsumption relations with other schemas. To add another subsumption relation to a schema, the user needs to select a subsuming schema on the ontology graph and press the Add ( ) button.

1.3      Define input / output data semantics

When a new schema is created, the I/O data descriptions inherited from its parent schema will be automatically included. A data description is composed of two parts: content and structure descriptions. The content description represents the semantics of data content and the structure description represents the semantics of data organization structure. The user can specialize the I/O data descriptions or add more descriptions.

To specialize the content or structure semantics of a data description, the user needs to select the corresponding cell in the I/O description table, choose a specialized content or structure schema from the data ontology graph, and press the Enter ( ) button. To check the position of the currently assigned content or structure schema in its ontology graph, users can double-click on its table cell. Then, the data ontology graph will be expanded and scrolled to show the corresponding node, and its schema properties will be displayed at the right-side panel.

To add more data descriptions, users can press the Add ( ) button. Then, an empty row will be added to the I/O data description table, and users can fill it with content and/or structure semantic descriptions as explained above. If the user presses the Add button after highlighting an existing row, the new data item will be added below the highlighted row. Otherwise, the new data item will be added at the top of the table.

If there is no appropriate ontology to describe the semantics of the service I/O data, the user needs to create a new content and/or structure schema(s). The  way to create a new data schema is similar to the way of creating a functional schema. Please refer to 2.1.1 and 2.1.2 for details.

2           Classify a service instance

After defining required functional and data schemas, users can classify a service instance under the functional ontology. To do this, users need to press the New Instance ( ) button after selecting an appropriate ontology on the service graph. Then, an editor (instance description editor) to describe the service instance will be displayed at the right-side panel. At the instance description editor, the user needs to specify the URI, a textual description, the name of the Job Request Entry for invoking the service, and syntactic descriptions of I/O data.

Syntactic information of data items can be described by their data instances and ordering between them. To specify data instance, the user needs to select the syntax column of a data item, select an appropriate data instance, and press the Enter ( ) button. The order between data items in the I/O data table must be matched against the order of parameters of the Job Request Entry. To reorder the data items of a service instance, the user can use the Move Up ( ) and Move Down ( ) buttons after highlighting the data item to relocate.

If there is no appropriate data instance to specify syntax of a data item, a new data instance that can syntactically represent the data item needs to be implemented and classified. To classify a data instance, users need to locate the appropriate ontology node on the data graph and press the New Instance ( ) button. Then, the user can edit the properties of the instance such as the URI, textual description, and Java class name. If the data instance has both content and structure semantics, it should be classified in the content ontology graph and its structure semantics can be specified by selecting an appropriate structure schema from the Organization Structure graph and press the Enter ( ) button beside the Structure Schema field.

3           Save and load a resource description

The component semantic description can be save to an XML file. If the user selects Save menu in the File main menu, the description will be saved to the default metadata file GeoWorlds/Client/GeoWorldsResource.xml. A new file path can be specified by selecting Save As menu in the File main menu.

Saved resource description can be loaded back to the resource description tool by selecting Open menu in the File main menu.

(C) Copyright 1998-2003 USC Information Sciences Institute. All Rights Reserved.
For problems or questions regarding this web contact [[email protected]].
Last updated: January 14, 2003 .