ISPRS, Vol.34, Part 2W2, “Dynamic and Multi-Dimensional GIS”, Bangkok, May 23-25, 2001
Service-ID
Location
Páramete
& Type
• Type
Input
data
schema
table
Output
data
schema
table
Data Schema Table:
Field-ID
Type
Available
operators
Constrains
Type
Table 1 Data model for services
Data service is specified by its identification, location of the
service provider (i.e. URL), parameters available to the service,
category of the service and the data set. Function service uses
a similar data structure as data service, with an additional
column that specify the schema of its output data. Data schema
table specifies the data input to a service, this is particularized
by operators and corresponding constraints. A single data set
corresponds to a record in the table. The table is used for the
input data to a data service and for both input and output data
for function service as well.
6. APPLICATION EXAMPLES
This section gives an application example to illustrate the
operation of the proposed Spatial Information Marketplace.
Supposes that there are two data service providers and one
function service providers in the Marketplace. Information of the
catalog is shown as below:
Service Table (Data service):
Service-ID
Location
Parameter &
Type
Type
Data
schema
table
A
http://a.com
(Return, string)
Data
1
B
http://b.com
(Return, string)
Data
2
Service Table (Function service)
Serv
ice-l
D
Locati
on
Parameter,
& Type
Type
Input
data
schema
table
Output
data
schema
table
C
http://
c.com
(Color, Look
up Table)
eMapping
3
4
Data Schema Table 1:
Field-ID
Type
Available
operators
Constrains Type
X
Real
Equal, <, >
10000-30000
Y
Real
Equal, <. >
10000-30000
Zoom level
Integer
Equal
1-8
Language
String
Equal
Chinese /
English
Data Schema Table 2.
Field-ID
Type
Available
operators
Constrains Type
Address
String
Equal
No + Street +
Building Name
Language
String
Equal
Chi / Eng
Zoom level
Integer
Equal
1 - 10
Data Schema Table 3:
Field-ID
Type
Available
operators
Constrains Type
~X1
Real
Equal
10000 - 30000
Y1
Real
Equal
10000-30000
X2
Real
Equal
10000-30000
Y2
Real
Equal
10000 - 30000
Land Use
String
Equal
A, B, C, D, E, All
Data Schema Table 4:
Field-ID
Type
Available
operators
Constrains Type
Filename
Image
NA
NA
Suppose the requesting language is implemented using Active
Server Page (ASP) in the Marketplace. Structure of the RSL is
known as:
Data service:
http://sericeprovide.asp?”Service=”&”Name of Service
provider” & “Data” & “Operantl” & ”Operator1” & “Constraint
1" & “Parameter” & “Parameter Type 1” & “Value 1”
Service, Data and Parameter are the keywords to delimitate the
three components of a statement of the language. Within the
component of Data, each filtering element is grouped by three
elements: “Operant", "Operator", and “Constraint”.
Function service:
http://sericeprovide.asp?”Service=”&”Name of Service
provider” & “Data" & “Operantl” & "Operatorl” & “Constraint
1” & “Output Data” & “Operantl” & "Operatorl” &
“Constraint 1” “Parameter" & “Parameter Type 1” & “Value
1”
There are three requests to the services and the corresponding
RSL statements are:
Request 1:
Retrieve the map at X=1000 and Y=2000, zoom level at 5,
language is English, and return the map in Shape file format.
Requesting language 1:
http://a.com/mapquervservice.asp?”Service-’&’’a”&"Data”&’’X"
&”-’&”1000”&"Y”&’’=’’&”2000”&”ZoornLever&’’=”5”&”Lanquaqe”
&"="Enqlish"&”Parameter”&”Retum”="Shape”
Request 2:
Retrieve the map with address “123, ABC Street”, language is
English, and return the map in JPEG format.
Requesting language 2:
http://b.com/mapquervservice.asp?”Service=”&"b"&”Data”&”Ad
dress”&"="&"123 ABC Street”&”Lanquaqe”&"=”&”English”&’’P
arameter”&”Return”&’-"&”jpq”
Request 3
Product the Land Use map for area within (1000, 1000) and
(2000,2000). Draw the map based on a lookup table
(lookuptable.txt). Return the map as “landuse.jpg”
Requesting language 3:
http://c.com/functionservice. asp?Service=”c’’&”lnputData”&"X1
1000”&“Y1 ”&"="&”1000''&“X2"&"="&”2000”&“X2"&”=”&”
2000"&”Landuse”&”=”&"AH"&"Parameter”&"Color”&"=”&“lookup
table.txt”&”OutputData"&"landuse.jpg”
7. CONCLUSIONS
Spatial Information Marketplaces will be the most popular way
of using GIServices in near future. The approach not only
provides actual economic benefits to both users and vendors of
GIS, but also follows the mainstream in information technology
(IT). This paper describes the architecture for the marketplace,
as well as a data model and requesting language for generic
services. The authors concludes the paper by giving key
technical barriers in the development of the Spatial Information
Marketplaces:
• Multiple server structure: Current standard in geographic
information interoperation is not comprehensive enough
for the use in the Spatial Information Marketplace. For
example, OGC is now investigating the cascading server
for multiple WMS server structure, this approach is not