Intelligent Agents Characteristics

Report on Intelligent Agents

Methodology

The research for this report was done using the various books from the library, Internet sites and computer magazines.

Introduction

The uses of intelligent agents within the home and in businesses have increased dramatically over the years. Intelligent agents are a part of a program that carries out a task unsupervised and applies some degree of intelligence to the task.

New and improved intelligent agents are constantly being designed and produced to carry out numerous repetitive and predictable tasks.

One definition of an intelligent agent, as described by states “an intelligent agent perceives its environment via sensors and acts rationally upon that environment with its effectors. Hence, an intelligent agent gets percepts one at a time and maps this percept into actions”

In his book, Essentials of Management Information Systems, Kenneth Laudon describes an intelligent agent as being “software programs that work in the background to carry out specific, repetitive and predictable tasks for an individual user business or software application”.

The design and production of an intelligent agent has to take into consideration numerous factors. This report takes a look at these considerations and factors and provides and insight into how intelligent agents are influencing businesses and society as a whole.

Findings

  • Agent Characteristics

Intelligent agents have four main characteristics:

“An agent is a computer software system whose characteristics are situatedness, autonomy, adapitvity and sociability.”

  • Situatedness

When an Agent receives some form of sensory input from its environment, it then performs some actions that change its environment in some way.

  • Autonomy

This agent characteristic means that an agent is able to act without direct intervention from humans or other agents. This type of agent has almost complete control over it own actions and internal state.

  • Adapitvity

This agent characteristic means that it is capable of reacting flexibly to changes within its environment. It is able to accept goal directed initiatives when appropriate and is also capable of learning from it’s own experiences, environment and interaction with others.

  • Sociability

This type of characteristic means that the agent is capable of interacting in a peer-to-peer manner with other agents or humans.

  • Design Considerations

One of the most important aspects of intelligent agents is the design of the actual agent. The agent needs to be able to fulfil the tasks that are required from it, i.e. to achieve its goals.

There are four main aspects that need to be taken into consideration when designing an intelligent agent.

  • Percepts

This is the information that the agent receives

  • Actions

This is what the agent needs to do or can do to achieve its objectives.

  • Goals

This is the factor that the agent is trying to achieve

  • Environment

The final aspect is the environment in which the agent will be working in. The environment in which the agent performs is probably the most important aspect that needs to be considered as this affects the outcome of the percepts, actions and goals.

  • Different Approaches to Agent Design
Read also  Effective Leadership In the NHS

From the agent characteristics and design considerations come four different approaches to intelligent agent design.

  • Simple Reflex Agents

Simple reflex agents are the most basic form of intelligent agent. They are simple minded, direct connections between percepts and actions.

  • Reflex Agents with Internal State

Reflex agents with internal state are similar to the Simple reflex agents except they remember the state of the environment as contained in earlier percepts.

As the agents’ sensors do not provide a detailed account of the environment at each input, a perception of the environment is captured over a period of time that provided further information to the agent and enables it to provide better results.

  • Goal Based Agents

For a goal based agent, the agent must know more than the current state of the environment, they must know the full requirements of the goal that they are required to perform.

The goal based agent combines the information of the goal with possible actions that will achieve that goal. This may cause the agent to take longer sequences of possible actions before deciding on the right course of action and whether the goal has been achieved.

Goal Based agents also take the future into consideration.

3.4 Utility Based Agents

Utility-based agents are the ultimate form of intelligent agent and are an extension of the goal-based agent.

Utility agents consider degrees of utility and try to maximise their own potential. Utility functions allow the agent to identify conflicting or alternative goals and decisions.

The likelihood of success and importance of the goal can also be compared and evaluated by the utility agent; the agent would then execute appropriate action to ensure the best option was selected.

  • Distinction between Environments
  • Accessible Vs Inaccessible

This environment is concerned with whether the agent is able to see the exact state of the environment. If the agents sensors are able to provide it with complete access to the state of the environment needed to choose an action, then the environment is accessible otherwise it is inaccessible.

  • Deterministic Vs Nondeterministic

In a deterministic environment, the use of the same actions will produce the same outcome every time the process or situation is repeated.

But in a Nondeterministic environment the final outcome of the process or situation will be different every time.

  • Episodic Vs Nonepisodic

In an episodic environment, subsequent episodes do not depend on the actions that occurred in previous episodes, agents in this type of environment don’t need to plan ahead.

Read also  Commitment To The Company

Agents in Nonepisodic environments do tend to depend on previous actions that occurred in previous episodes and do tend to have to plan ahead.

4.4 Static Vs Dynamic

A static environment is an environment that doesn’t change and the time taken for an agent to process an action doesn’t matter, as the environment will remain the same.

A dynamic environment changes continuously and an intelligent agent has to be able to process actions quickly before the environment changes.

  • Discrete Vs Continuous

In a discrete environment the number of distinct percepts and actions that an agent will receive is limited to a set amount, but in a continuous environment the percepts and actions are unlimited.

Intelligent agents that are in an accessible or deterministic environment do not need to deal with uncertainty.

Inaccessible, Nondeterministic, Nonepisodic, Dynamic, and Continuous environments are the most challenging, as these are unpredictable environments, the other environments are more stable and less volatile.

  • The use of Intelligent Agents in Businesses

The use of intelligent agents has increased dramatically over the past 5 year, though the majority of people wouldn’t think twice about the process that they do that make their lives easier.

During the1990’s, A.D.Little, a management consultant, estimated that by the year 2000, 15 to 20 percent of all computer applications would contain intelligent agents. Though this figure has increased a substantial amount with the development of new technologies.

The use of intelligent can agents can be seen in every industry sector within the UK and all over the world.

The most common and widely used agents are found within two main areas. The first is within office type environments in which computers are used and the second is on the World Wide Web.

In the first issue, computer software is the main reason for agents being used so widely:

  • Operating systems use agents to add email and dial up networking account, do group management, add/remove programs and devices and monitor licences.
  • Spreadsheet agents offer suggestions for improvement and can also tutor novice users.
  • Software development agents assist in routine activities such as data filtering.

The second and area and probably the largest is the Internet. The internet uses a variety of different types of agents to help the user find what they are looking for. They include:

  • Search engines – improve your information retrieval on the Internet
  • Web mastering Agents – these agents make it easy to manage a web site
  • Web Agents – These agents improve the users browsing experience.
  • Monitoring Agents – These agents monitor web sites or specific themes you are interested in.
  • Shopbots – These agents allow you to compare prices on Internet.
  • Virtual Assistants – these include virtual pets and desktop assistants.
Read also  Management Essays - Bread Bakery

Intelligent agents can be seen in a wide variety of situations, the table in point 5.1 provides more examples of what agents are capable of.

Though agents are making life easier, it is also reducing the amount of employees needed to do the job. An example of this would be a car-manufacturing factory. During the 1980’s thousands of people were employed to make the vehicles but since the introduction of machinery that contains an intelligent agent to do the repetitive work, numerous people were made redundant and the positions in which 4 – 5 people did the job has been reduced to one or two men and the machine.

  • An intelligent agent is something that perceives and acts in an environments.
  • An ideal agent is one that always takes the action that is expected to maximise its performance.
  • There are a variety of basic agent program designs. The designs vary in efficiency, compactness and flexibility. The appropriate design of the agent program depends on the percepts, actions, goals and environment.
  • Some environments are more demanding than others.
  • Reflex agents respond immediately to percepts, goal based agents act so that they can achieve their goals and utility-based agents try to maximise their own ‘happiness’.
  • The use of intelligence agents has increased beyond the expectations of experts within the management and information technology fields. The use of intelligent agents are being seen in a wide cross section of businesses whether it be in machinery and equipment or within the software programs that they have in their computers and networks.
  • The most widely used forms of intelligent agents are found on the Internet, they are mainly used within search engines.

Recommendations

  • Intelligent agents are the basis of artificial intelligence; there are considerable ongoing researches into the field, with many exciting possibilities.
  • “Agents are here to stay, not least because of their diversity, their wide range of applications and the broad spectrum of companies investing in them. As we move further and further into the information age, any information-based organisation which does not invest in agent technology may be committing commercial hara-kiri.” –

Bibliography

Stuart Russell and Peter Norvig 1995,Artificial intelligence: A modern Approach, Prentice Hall

www-cdr.Stanford.edu/nextlink/expert.html [Accessed 13 October 2002]

Intelligent Agents on the internet: fact, Fiction and Forecast, Oren Etzioni and Daniel S. Weld – www.computer.org/intelligent/ex1995/x4044abs.htm [Accessed 16th October 2002]

www.hermans.org/agents/index.html [Accessed 16th October 2002]

Intelligent agents likely to cut jobs, alter offices, Kile Martz

www.bizjournals.com/kansascity/stories/1998/08/03focus5.html [Accessed 16th October 2002]

http://www.agentland.com/

http://sbm-connect.tees.ac.uk/ebuscon/Presentations%20(PPP%20&%20PDF)%20Files/Chapter%2011%20Intelligent%20Systems%20in%20Business.pdf [Accessed 13th October 2002]

Turban, McLean and Wetherbe, 1996, First Edition, Information technology for managers – Improving Quality and Productivity

Laudon K. Essentials of Management Information Systems, 2002, Fifth Edition, Prentice Hall

Order Now

Order Now

Type of Paper
Subject
Deadline
Number of Pages
(275 words)