Demo Instruction

This demo instruction is prepared as a supplement to the submitted paper .

Two typical examples are demonstrated using the OWLSManager, a component for OWL-S Files Management, which can deploy and undeploy OWL-S files into the knowledge base.
   The first example is used to demonstrate the subsumption reasoning to get the semantically matched Web services in the service composition process. Different geo data types(high-level geo data product) represented as ontology entities in the query XML can be produced based on an automatically and dynamically generated service chain whenever the CSW service is available and the service and data registration can be searched using the CSW service.
   Furthermore, the second example demonstrates the mediated RDF structure as the metadata relay structure strategy in the service composition process to facilitate the metadata tracking and ECA rule usage in the OWL-S preconditions satisfiability. The taxonomy classify process is necessary to check the OWL-S precondition, for example, the precondition on the data format in the http://www.laits.gmu.edu/geo/ontology/owls/ap/v2/slope_precondition.owl needs to verify whether SupportedFileFormats(Collection of ...#GeoTIFF and ...#NetCDF) is satisfied by the input data's RDF structure element "&iso19115;#distributionFormat".

Demo steps: Demo Flash

1. Users first go to Service Composition page through the left navigation menu.
2. To view the landslide risk sample(Demo example 1), users click the link of OWL-S Manager for Web Service Discovery and Chaining.
To view the metadata tracking and ECA rule demo(Demo example 2), users click the entry link of Metadata Tracking.
3. Users then click Service Chaining and Execution link on the left navigation menu of the opened Web page. In the newly opened page on the right, click the submit button. You will see the running result.

Notes: for the Demo 1, you have to choose the "SUBSUME" Option of DataType Match to get the available landslide risk data product. The requested data type in the <Ontology> tag of default request xml is ".../GeoDataType.owl#Terrain_Slope". To view the landslide risk data product, you have to change it to the ".../GeoDataType.owl#Landslide_Susceptibility".

Sample results for explanations:

Geospatial data, Web services and Web applications are distributed on the Geobrain, Laits and Data servers. Geobrain server is located at NASA GSFC. Laits and Data server are located at CSISS, GMU.

Service	Description
Landslide Susceptibility	The computational model for landslide susceptibility in this service takes into consideration the factors of terrain slope, terrain aspect, land cover types, and vegetation conditions (through the Normalized Difference Vegetation Index, or NDVI) by assigning each a weighting factor and then doing the map algebra computation.
Slope	Computes the terrain slope based on Digital Elevation Model data(DEM).
Slope Aspect	Generates the terrain aspect based on DEM data.
ETM NDVI	Calculates ETM(Landsat Enhanced Thematic Mapper imagery) NDVI based on the NIR image and red image.
OGC WICS (Web Image Classification Service)	Performs the image classification functions (supervised) that can generate the land cover types.
OGC WCS (Web Coverage Service)	Provides the available geospatial data in the data archives
OGC WCTS (Web Coordinate Transformation Service)	Performs the reprojection computation that can transform the data from one spatial projection to another spatial projection.
DFTS(Data Format Translation Service)	Performs the reformating computation that can transform the data from one file format to another file format.
OGC CSW(Catalogue Service for the Web)	Web-based geospatial catalog service for publication, discovery, and access of geospatial data and service.

Demo 1 (average running time 27s):

view 1
view 2

Explanations as follows:

Request XML: Ontology entity(i.e. "Data Type") and the geospatial constraints(i.e. metadata requirement).
Physical Model: A link to the graph structure for the produced service chain.
CSW Data Query Result: The available data queried from CSW as the input to the produced service chain.
Composite Process: OWL-S composite process for the produced service chain.
OWL-S Input Values: They can be used to invoke the composite process.
Question Answer: Landslide susceptibility data product resulting from the service composition.
User can copy the linkage contained in the question answer to the Data File URL for HDF Viewer textbox and click the view button, the HDF data file will be transformed to PNG image for your convenient check in the Web browser.
Landslide Susceptibility Image resulting from the execution of the service chain for the requested XML.

Demo 2(average running time 65s):

View 1
View 2

Explanations as follows:

Request XML: Ontology entity(i.e. "Data Type") and the geospatial constraints(i.e. metadata requirement).
Physical Model: A link to the graph structure for the produced service chain.
CSW Data Query Result: The available data queried from CSW as the input to the produced service chain.
Composite Process: OWL-S composite process for the produced service chain.
OWL-S Input Values: They can be used to invoke the composite process.
Question Answer: Slope data product resulting from the service composition.
The final data product is a geotiff format thus it can be supported by many popular picture viewer software.

Comparison:

Compared with Demo 1, demo 2 takes more time. It is mainly due to two reasons:
(1)Although Demo 1 involves much more services, it is a parallel computation.
(2)OWL-S files in Demo 2 are assigned with OWL-S preconditions. It shows that precondition check and SWRL processing consumes additional time, especially when a taxonomy classify process is invovled in a comparatively large knowledge base. This implies that the performance of our current implementation could be improved further by optimizing the reasoning engine.