|
Home
Synopsis
Research
Papers |
Conceptual framework and code-structure generation methods
and tools for trade agent technology
Based on the results of work units 3 and 4, a conceptual
framework will be defined for agent technology with respect to the kind
of problems that will benefit of the application of this technology.
The framework supports the relation between conceptual
characterisations of problem domains and conceptual characterisations of
agent architectures. Special attention will also be given to security requirements
and properties.
The framework allows selection and preparation of the implementation
of an appropriate agent technology, based on the characteristics of the
problem that has to be solved. Also, the framework provides a standardised
manner to express the conceptual model behind an agent architecture, its
properties, required environment assumptions, and the
assumptions on available types of domain knowledge. The framework also
provides a standardised manner to express different problem types in terms
of domain structure, domain constraints and assumptions, and requirements
on functionality and security.
Subsequently, by using this framework as a basis, methods
and tools for selecting appropriate agent technologies and generating code
structure for trade agent technology are developed. (The type of generation
of code structure is comparable to that of the tool Rational Rose.)
Finally, the methods and tools are fine-tuned based on
the continuing conclusions from the experiments, applications, and usage
in work unit 4 and work unit
6.
-
Currently available deliverables
[1D5.1] Internal
Agent Architecture (restricted access only)
This report is the result of Task T 5.1. The goal of
this activity is stated in the projectplan as follows: "The goal of this
activity is to develop an architecture that allows a developer to take
an implementation of an agent core, plug it into an agent of an agent technology
provider and integrate it with its communication infrastructure with the
least effort possible. The architecture ideally provides standard facilities
for receiving messages and notifications of changes in the environment,
the execution of actions and an interface for a Reasoner Component. Furthermore,
the architecture enables each autonomous agent to run in its own thread
of control".
|
