The development of software agents is
a recent trend in both distributed computing research and geographic
information science (GIScience).
Software agents can help users access distributed data objects and
programs on heterogeneous geographic information systems (GIS) platforms
by interpreting, filtering, and converting information automatically.
This project aims to facilitate the understanding of the role of
software agents in distributed geographic information services (GIServices)
and introduce the research of software agents to the GIScience
community.
This page will present a introduction to the project, including:
- Project's objectives.
- Project's tasks.
1) Design of geoagent model and geoagent
communication protocol.
2) Development of a prototype for software agents.
3) Evaluation of the prototype model by implementation of real geodata and
GIS operations.
For further understanding the basic information of GIS and agent,
please refer to the relevant pages.
Objectives of the Project Research
This research will design a software agent communication
mechanism to facilitate the integration of heterogeneous geospatial
data, GIS programs, and modeling procedures in distributed network
environment. The Unified Modeling Language (UML) will be used to design
a software agent model. A geoagent communication protocol (GCP) will be
developed by using Extensible Markup Language (XML) and Knowledge Query
and Manipulation Language (KQML) to examine the communication
requirements of distributed GIServices. A prototype will be developed on
two commercial GIS software packages, ARC/INFO and ArcIMS. For further
background information, please refer to the Agent and GIS pages.
[TOP]
Tasks of the Project Research
Task 1: The design of software agent model and geoagent
communication protocol (GCP) for distributed GIScience
The design of software agent model will include the formalization of
agent object classes, interactions and responsibility for different GIS
tasks, such as the communications among different machines, the
integratiion of GIS components, and the searching for requested geodata
objects, etc. This research will focus on the design of object classes
of software agent model by using the Unified Modeling Language to design
the object model for software agents. Three class objects of software
agents, geodata agents, component agents, and machine agents, for
distributed GIServices will be developed and specified based on their
functionality. The adoption of GCP will facilitate the communication
between software agents with heteogeneous distributed databases and GIS
components in dynamic network environment. GCP development will adopt
the concepts and framework of KQML to the development of GIService
applications. Four types of agent communication mechanisms,
agent-to-agent, agent-to-database, agent-to-program, and
agent-to-machine will be considered. Each type of agent communication
will also include two sub-types of communication approach, local and
remote.
Task 2: The development of a prototype for software agents by
customizing current GIS packages, ARC/INFO and ArcIMS
In order to test the feasibility of software agent model and GCP,
this research will develop a prototype by adopting current available GIS
software and technology. Two GIS software packages, ARC/INFO and ArcIMS
will be customized taking advantage their current network communication
architecture. ArcIMS provides a ready-to-use, low-level network
communication capability, which can be utilized directly for GCP and
software agent model. ARC/INFO's GIS functions can be customized by its
ArcObject development tools, to create a remote invocation method for
GIS functions and database connectivity. The advantages of adopting
these two GIS commercial packages are to allow this research to focus on
high-level, semantic design and application services instead of spending
too much effort on low-level network programming and client-server GIS
functions.
Task 3: The evaluation of the prototype model by conducting
several GIS scenarios with the actual implementation of real geodata and
GIS operations.
The prototype of software agent model and GCP will be applied to
several hypothetical GIS scenarios, such as spatial decision support
systems, data mining, and mobile device integration with actual data
implementation and GIS operations. The design of scenarios will focus on
such issues as distributed GIS procedures, automatic data conversion,
the efficient use of computing resources, cross-platform applications,
real-time data update.
|