Multi Agent Systems -an introduction-

1
Multi Agent Systems-an introduction-
2
Content

• What is an agent?
• Communication
• Ontologies
• Mobility
• Mutability
• Applications

3
What is an agent?
4
Definition

• An autonomous agent is a system situated within
  and part of an environment that senses that
  environment and acts on it, over time, in pursuit
  of its own agenda and so as to effect what it
  senses in the future.
• Franklin and Graesser
• Is it an agent or just a
  program?
• Simpler agents are autonomous programs

5
Coming from AI

• Expert systems did not prove commercially
  successful too ambitious
• Agents provide a way to use limited amount of
  intelligence to achieve useful goals
• Russell Norvig Artificial Intelligence, a
  Modern Approach
• Modern Agent based

6
Coming from OOP

• Allows us to use more complex entities than
  traditional OOP
• Active components, not handled in OOP
• OOP definition of agent object with a thread
• Simplistic, but it can be used as a marketing
  term
• Software engineering techniques need to be updated

7
Coming from Client-Server computing

• Client originates requests, server responds
• The model is too simple, many times we need
  active components, e.g. for monitoring
• Enterprise Java Beans actively discourage beans
  with threads
• They have their reasons, but limits functionality
• There are workarounds but will need to be handled
  at the standards level

8
People also come from

• Business
• Agents seen as economic entities
• Sociology
• Social models of agent
• Agent models of the society
• Phychology
• Behavior patterns
• User Interaction Design
• Philosophy
• Lots of sources of confusion.

9
Content of this course
10
EEL 5937 Practice and theory of multi-agent
systems

• Practice We will learn how to write agents
• Theory We will learn the basics behind the
  agent systems.
• Topics covered
• Behaviors, strategies, reactive and planned
  behavior
• Belief-Desire-Intention models
• Communication, messaging, speech act theory
• Mobility
• Mutability
• Knowledge representation, ontologies
• Negotiation models
• All topics covered will be used in some of the
  projects

11
Contact info

• Instructor Lotzi Bölöni ENG 444
• E-mail Lboloni_at_cpe.ucf.edu
• Webpage http//www.cs.ucf.edu/lboloni/
• Course webpage
• http//www.cs.ucf.edu/courses/eel5937-multiagent/
• FEEDS Video stream
• Link from the course webpage

12
Grading

• Homeworks reading 20
• Homeworks small assignments to get you up to
  speed for the project.
• Reading assignments interesting, famous,
  controversial etc. papers in the domain.
• Final exam (written, open book) 20
• Project 60
• Additional credit core development for the Bond
  system (up to 20)

13
The project

• Groups of 2-4 people.
• Application of multi agent systems.
• Proposed configuration Bond agent framework
• Bond libraries and tools
• Jade agent framework
• Protégé ontology system
• Java 1.4
• List of projects (growing)
• http//www.cs.ucf.edu/courses/eel5937-multiagent/
• You are encouraged to propose your own project.
• Teams should schedule a design meeting with the
  prof.