Home Articles A study on the component classification in GIS (Geographic Information System)

A study on the component classification in GIS (Geographic Information System)

Yun-Won Jo, Myung-Hee Jo, Dong-Ho Shin
Department of Urban Information Engineering, Kyungil University
33 Buho-ri, Hayang-up, Kyungsan-si, Kyungsan bukdo, 712-701, Korea
Tel : +82-53-850-7312,
Fax : +82-53-854-1272
Email :[email protected], [email protected], [email protected]

 

1. Introduction
Recently many software engineers have focused on component research to increase the reusability and interoperability. Especially, the research of common repository system for architecture technology, management, and development has been regarded as a big issue because it can save time and manpower by accessing to exactly what system developers are looking for.

System developers usually tend to spend much of their time and effort to find out certain functionality for the implementation of their preferred style. Also, they want to obtain guaranteed component for improving their system productivity and quality at the same time. However, as you know, it is very difficult to identify certain components in a repository because of the lack of quality software components and the inability of developers to efficiently find them. Here, in order to solve this, establishing fully repository architecture in the Web for component sharing and circulation should be encouraged in a proper way.

In order to implement GIS (Geographic Information System) more efficiently in the view of cost and time, GIS developers started to consider the concept of "Component GIS". It is mainly focused on reusability and interoperability because most GIS projects have its certain application such as MIS(Marketing Information System), ITS(Intelligent Information System), LIS(Land Information System), DCS(Disaster Control System), and FMS(Facility Management System) and each application needs its common functionalities such as mapping and query or its certain functionality such as 3D viewer and GPS data processing. Therefore, if there exists a universal repository storing GIS components and system designers or developers know where desired component blocks are located in real time, GIS developers can easily select their desired component then modify or composite their system by using them.

In this paper, the metadata, which describe component, are the key to the functionality of the repository. Therefore, metadata, especially describe GIS components the most, should be first defined then the architecture of a component repository should be constructed.
 


 

 

 
2. Research Background in OGC (OpenGIS Consortium)
What is the first thing that GIS application developers consider when they are developing the application system? As you know, just like any other system developers they must maintain its interoperability and reusability by using previous component.

In this paper, the OGC (OpenGIS Consortium) is more focused as related work because it has dedicated much time and effort towards solving the interoperability issues outlined previously. The OGC is "the full integration of geo-spatial data and geo-processing resources into mainstream computing and the widespread use of interoperable geo-processing software and geo-data products throughout the information infrastructure". In order to facilitate this, its working groups have developed abstract specifications and implementation specifications for its two central technology themes of sharing geo-spatial and providing geo-spatial services [1].

The Abstract Specifications [2]
The abstract specification documents provide the theoretical background for the implementation specifications. The Open GIS abstract specification documents are composed of two models: 1) the essential model that describes a conceptual link between the software system and the real world, and 2) the abstract Model that describes how the eventual software system should work in an implementation neutral manner.

Following lists briefly shows each category described in the Open GIS abstract specifications; Overview, Feature Geometry, Spatial Reference System, Location Geometry, Stored Functions and Interpolation, The OpenGIS Feature, The Coverage Type, Earth Imagery, Relationships between Features, Quality, Feature Collection, Metadata, The OpenGIS Service Architecture, Catalog Services, Semantics and Information, image Exploit Services, Image Coordinate Transformation Service.

The Implementation Specifications
The Implementation Specifications documents are a set of specifications that contain guidelines for implementing Open GIS applications or components.

Following lists briefly shows each category described in the Open GIS Implementation Specifications; Simple Features Specification, Catalog Services Interface Implementation Specification, Grid Coverage Implementation Specification, Coordinate Transformation Services Implementation Specification, Web map Server Interface Implementation Specification, Geography Makeup Language (GML) Implementation Specification.
 


 

 
3. GIS Component Classification using Metadata
In order to classify GIS component, there are two views of classification; 1) The functional component classification [3], 2) The non-functional component classification.

The functional component classification in GIS considers following 4 layers. First of all, Infrastructure Component, which describes general environment for developing component and executing such as various platforms, operation systems, and networks environment. Second, Data Provider Component, which has purpose to acquire interoperability among them constructed under heterogeneous environment, data provider component should be constructed. Third, Core / Base Component, which can be used as the kernel of GIS by developing certain GIS application software. There exist core components to perform general GIS functions such as Data Provider, Map Display, and Attribute Display also exists a base component to perform certain GIS functions such Network Analysis, 3D Analysis, and Authority. Forth, GIS Component, which shows several GIS domain such as MIS (Marketing Information System), FMS (Facility Management System), DCS(Disaster Control System), ITS(Intelligent Transformation System), UIS (Urban Information System), and LIS(Land Information System).

Even though there are certain components having reusability and good quality, the rate of their reusability may be low because they have difficulty to access in a repository. It is essential to express the non-functional element to identify components. Here, two big non-functional classification categories are discussed such as GIS content-dependent metadata and GIS contents-independent metadata in Fig. 1.
 

Figure 1 Non-Functional Classification of GIS Component
GIS content-dependent metadata: As you see in the item name, when metadata is associated with the original GIS domain itself, we call it GIS Content-dependent metadata. Especially, considering GIS component, there are three small categories related to it; 1) Identifier Attribute, 2) Service Attribute, 3) Support Data Attribute.

On the other hand, GIS content-independent metadata: it doesn't depend on the GIS domain. These kinds of metadata can be derived independently from the content of the GIS. There are seven small categories; 1) description Attribute, 2) Environment Attribute, 3) Development Attribute, 4) Restriction Attribute, 5) Supplier Attribute, 6) Quality Attribute, 7) Special Mention Attribute.

In Fig. 2, the needs of individual GIS system developers tend to limit to what kinds of component they are accessing to on Web. System developers can obtain certain functionality presented to them in a way suiting their preferred style for its implementation. As the result, they can save their time, effort, and cost through on-line adaptation of GIS component in a repository. At this time, Metadata query broker is working as a brokerage mechanism supplied in the repository for the retrieval of various types component. As stated before the original purpose of these metadata query broker is to search the component that is usually associated with implementation of GIS application. Especially, to avoid storing multiple copies in a repository, each component should be registered through CUID.
 

Figure 2 On-line Adaptation of GIS Component


 

 
4. Prototype of GIS Component Retrieval System on line
In this paper, the prototype of GIS component retrieval system is presented to register new component and manage them and retrieval related component effectively. Fig. 3 shows the prototype of Web based GIS component retrieval system. To operate this retrieval system, the non-functional GIS component classification, metadata of each GIS, is used.
 

Figure 3 The Prototype of GIS Component Retrieval System
5. Conclusion
In order to increase reusability higher, a large sized collection of reusable software is a necessity. Recently, most research efforts have been focused on not only the construction of repository that contains a large number of components but also define metadata well to retrieval.

In this paper, to operate GIS Component Retrieval System, two types non-functional component classifications in GIS are proposed; 1) GIS content-dependent metadata, 2) GIS content-independent metadata. Also, the detail metadata item for each are defined.

In the future, in order to increase component reusability in GIS domain, these should be consideration for GIS component repository management system using not only metadata described here but also further extended metadata.

Reference

  • Richard Michael Preston, "RADGIS-An improved architecture for runtime-extensible, distributed GIS applications", PhD Thesis, Rhodes University, 2001.
  • OpenGIS Consortium, The OpenGIS Guide, OpenGIS Consortium Inc, 1998.
  • E. J. Han, "Architecture Construction for For/With Component in GIS", The 2nd Annual International Conference on Computer and Information Science (ICIS '02), 2002.
  • H. K. Kim, E. J. Han, H. J. Shin and C. H. Kim, "Component Classification for CBD Repository Construction", Software Engineering Applied to Networking & Parallel/Distributed Computing, pp.483-493, 2000.
  • Y. W. Jo, M. H. Jo, S. S. Ahn, "Case Study of UML(Unified Modeling Language) Design for Web-based Forest Fire Hazard Index Presentation System", Journal of the Korean Association of Geographic Information Studies 2002, 2002.3, pp.58-68.