Agent technology
Agent technology
MAD
Agent
• Agents have an identity. A state and a behavior. State
(knowledge, goals, and responsibilities). Behavior (roles
that can be played action that can be performed, reaction
to events.
• The behavior of an agents is defined in term of how to
decide, what to do.
Continue …
• Agents can be considered as active objects.
• Agents unified modelling language (AUML) is now the
new language for designing and analyzing agents.
• Distributing computing language (CORBA, RMI), applets,
mobile object system, coordination mechanism language
can be used to implement agent -system.
DIFFERENCE BETWEEN
SOFTWARE AGENTS AND OBJECTS:
• Software agents embody a stronger notion of autonomy
than objects.
• Software agents are capable of flexible behavior.
• A multi -agents system is inherently multi-threaded.
FROM AI TO IA THROUGH
AGENTS:
• Artificial intelligence(AI): Aims to build system that can be
ultimately understand natural language.
• INTELLIGENT AGENTS (IA): Are system that are built in a
way that they can choose right action to perform at the
right moment in a very specific domain.
DISTRIBUTED COMPUTING
SYSTEM:
• Agents as part of distributed program running as a single
process.
• DISTRIBUTED AI SYSTEM: Agents is an active objects
+beliefs, desires, and intension.
AGENTS TYPES:
• Agents function and agents program
• AGENTS FUNCTION : Mapping a percept to
action.
• EXAMPLE: in agent function we plane about the output it
is not final that we give the output to the user not .
AGENT PROGRAM: Actual
implementation.
• EXAMPLE : our plane prepared and then implementation.
CONCLUSION:
• Agents represents concrete implementation of AI
principles, And advanced abstraction of OO techniques.