Topic 8 - School of Engineering and Information Technology

hurtpotatocreekΑσφάλεια

5 Νοε 2013 (πριν από 3 χρόνια και 7 μήνες)

76 εμφανίσεις

ICT619 Intelligent
Systems



Topic 8: Intelligent Agents

ICT619

2

Intelligent Agents


What is an intelligent agent?


Why intelligent agents?


What intelligent agents can do for us


Characteristics of a good agent


Types of agents


Building intelligent agents


Intelligent agents in E
-
Commerce


Intelligent agent design
-

state
-
of
-
the
-
art
and future


ICT619

3

What is an intelligent agent?


Underlying concept
-



An autonomous computational entity designed to perform a
specific task, without direct initiation and continuous monitoring on
part of the user


Emerged in the last 15 years or so


Distinct from conventional programs, in that it is
automatic


Additional properties:


Some level of intelligence (based on any AI technology from fixed
rules to learning engines) for decisions and/or adaptation to
environmental change


Acts reactively, but also proactively


Social ability
-

communicates with user, system, other agents as
required


Might cooperate with other agents to carry out complex tasks


Agents might move from one system to another to access remote
resources and/or meet other agents

ICT619

4

What is an intelligent agent? (cont’d)



Intelligent agents (also called “software agents”) do not
necessarily possess
all

these possible features



Wide range of variation in capabilities:



Some perform tasks individually while others are
cooperative



Some are mobile
-

able to move across a network,
others are not




Most communicate via coded messages or even
natural language, some don't communicate at all




Multiple agents

work in groups or swarms to solve
problems collectively, some work as individual units



Not all agents learn and adapt themselves



Robots are physically embodied agents

ICT619

5

Why intelligent agents?


More and more everyday tasks becoming computer
-
based


An increasing number of untrained users using computers


Current human
-
computer interfaces require users to initiate all
tasks and monitor them
-

manually



Intelligent agents engage in a cooperative process with the user to
leverage the effectiveness and efficiency of human
-
computer
interaction


Staggering growth in information availability


Intelligent agents can be a tool for relieving the user of this
information overload


Intelligent agents can act as personal assistants to the user to
manage information


Might one day take over routine tasks in personal management
such as appointments, meetings and travel arrangements

ICT619

6

What intelligent agents can do for us


Carry out tasks on the user’s behalf


Train or teach the user


Help different users collaborate


Monitor events and procedures



Specifically, intelligent agents can help us with


Information retrieval


Information filtering


Mail management


Recreational activities


selection of

books, music, holidays


Booking of meetings, hotels, tickets


ICT619

7

What intelligent agents can do for us
(cont’d)

Information filtering agent


One type is the selection of articles from a continuous
stream to suit particular user needs



User can create “news agents” and train them by giving
positive or negative feedback for articles recommended



The use of key words alone can be restrictive


Underlying semantics must be extracted for more
effectiveness


Eg VPOP Technologies' Newshub
-

an automated,
agent
-
based web news feeder service, which delivers
customised updates of stories from major news outlets
every 15 minutes

ICT619

8

What intelligent agents can do for us
(cont’d)

Electronic mail agent


Assist users with electronic mail


Learn to prioritize, delete, forward, sort and archive
mail messages on behalf of the user


May use intelligent system techniques like case
-
based
reasoning


Can associate a level of confidence with its action or
suggestion


Use of “do
-
it” and “tell
-
me” thresholds set by user


May involve multi
-
agent collaboration

ICT619

9

What intelligent agents can do for us
(cont’d)

Selection agents for entertainment


Conversational agents show potential for

becoming popular and commercially

successful eg Cybelle, ALICE




Use “social filtering”


correlation between different
users to make recommendations on books, CDs, films
etc.


So, if user
A

liked items
X

and
Y
, and user
B

liked item
X

and
Z
, then item
Z

may be recommended for user
A



Amazon.com has been using this system for years
-
>



Hi, I am Cybelle.

What is your name?

ICT619

10

ICT619

11

What intelligent agents can do for us
(cont’d)

Some other current and emerging applications of
intelligent agents:


air traffic control


air craft mission analysis


control of telecommunications and network systems


provision and monitoring of medical care


monitoring and control of industrial processes


on
-
line fault diagnosis and malfunction handling


supervision and control of manufacturing environments


transactions management in banks and insurance
companies


E
-
commerce, tourism

ICT619

12

Characteristics of a good agent

Action


Agent must be able to take some action and not just
provide advice


Present state of web technology limits capability of
Internet agents


-

still no standard interface for agents, but agent
communication languages such as ACL and KQML
might win out


As the Internet becomes more agent
-
friendly, more
capable agents will emerge


Autonomy


An agent can be much more useful if it can act
autonomously


The right level of autonomy for a task must be found

ICT619

13

Characteristics of a good agent
(cont.)


Communication


Must communicate well with the user


Should understand user’s goals, preferences and constraints


Useful communication requires shared knowledge on


language of communication


problem domain


Example Problem: Web search engines


accept key words and phrases (some knowledge of the
language)

but


understand nothing about the documents they retrieve (no
domain knowledge)


Solution: provision of a machine
-
readable
ontology



-

a definition of a body of knowledge including its


components and their relationships

ICT619

14

Characteristics of a good agent
(cont.)

Adaptation


Can gain user confidence by learning user preferences


ML techniques such as ANNS, GAs or CBR can be
used


Adapting to user preferences can be also achieved by
using data mining techniques such as clustering


Agent forms clusters of users with similar features


User's needs can then be anticipated by placing the
user in one of these clusters and analysing the cluster


Social problem solving method, similar to Amazon
recommendations

ICT619

15

Types of agents


Based on operational characteristics and
functional objectives:


Collaborative agents


Work together to

-

integrate information and

-

negotiate with other agents to resolve conflict

-

Provide solutions to inherently distributed problems,
e.g., air traffic control



Reactive agents


Act by stimulus
-
response to the current state of
the environment


Each reactive agent is simple and interacts
with others in a basic way

ICT619

16

Types of agents (cont’d)

Interface agents


Provide user support and assistance


Cooperate with user in accomplishing some task in an
application.


Interface agents learn:


by observing and imitating the user


through receiving feedback from the user


by receiving explicit instructions


by asking other agents for advice (from peers)


Examples:


Personal assistants performing information filtering,
email management.



ICT619

17

Types of agents (cont.)

Mobile agents


Programs that migrate from one machine to another.


Execute in a platform
-
independent execution environment, like
Java applets running on a Java virtual machine


Practical but non
-
functional advantages:


Reduced communication cost


Asynchronous computing (when you are not connected)

ICT619

18

Types of agents (cont.)

Two types of mobile agents:


One
-
hop mobile agents (migrates to one other
place)


Multi
-
hop mobile agents (roam the network
from place to place)



Example applications:


Distributed information retrieval


Telecommunication network routing

ICT619

19

Types of agents (cont.)

Information agents


Manage information


Manipulate or collate information from many distributed
sources.


Can be mobile or static.


Examples:


BargainFinder compares prices among Internet stores for
CDs


Jasper works on behalf of a user or community of users and
stores, retrieves and informs other agents of useful
information on the WWW


ICT619

20

Types of agents (cont.)

Multiple agent systems


Consist of collections, or swarms, of simple agents that
interact with each other and the problem environment


Can be mobile or static, same or different agents


Complex patterns of behaviour emerge from collective
interaction


Examples:


Swarm of bees finds an optimal location for the hive


xxxx


ICT619

21

ICT619

22

Building intelligent agents

Two main problems to overcome:



Competence



How do we build agents with the knowledge needed to decide


when to help the user


what to help the user with, and


how to help the user?


Trust



How to guarantee user comfort (and protection!) in

delegating tasks to the agent



Approaches to building agents

1.
User
-
programmed agents
-

write specialised scripts

2.
Knowledge
-
based agents

3.
Machine
-
learning approach

ICT619

23

Building intelligent agents (cont’d)


The main problem with
user
-
programmed

approach


-

requires high level of user competency


-

user must be able to


Recognise opportunity for employing an agent


Take initiative to create an agent


Impart specific knowledge to agent by codifying it in a
special language


Maintain agent’s knowledge by updating rule base with
time



The issue of trust is then reduced to users’ trust in
their own programming skills

ICT619

24

Building intelligent agents (cont.)


In the
knowledge
-
based approach,



The agent is supplied with knowledge about
the application and user



At run
-
time, agent uses the knowledge to
recognise user’s plans and find opportunities

to contribute to them



Example of knowledge
-
based agent: the
UCEgo
-

designed to help users solve
problems in using the UNIX operating system.


ICT619

25

Building intelligent agents (cont.)


Problems with knowledge
-
based approach
-



Both competence and trust are issues of concern



The problem of competence relates to the competence
of the knowledge engineer



Knowledge
-
base is fixed and cannot be customised to
specific user needs



User’s trust is affected as agent is programmed by
someone else

ICT619

26

Building agents


the machine
learning approach


Metaphor of a personal office assistant


Agents start with minimum knowledge and learn
from:

1.
Observation and imitation of user

2.
User feedback


direct, indirect

3.
Training by user

4.
Other agents



User can build up model of agent decision making


more trust


Agent capable of explanation

ICT619

27

Development of an agent through
learning

ICT619

28

Building agents


the machine
learning approach

Advantages:


Less work from end
-
user and developer


Agent customises to user/organisation
habits/preferences


Helps distribute know
-
how and competence

among different users


Some examples:


Agent for e
-
mail handling


Agent for meeting scheduling


Agent for electronic news filtering


Agent for recommending books, music


ICT619

29

Intelligent agents in E
-
commerce


Rapid growth continues in e
-
commerce


Information about products and vendors is easily
accessible


But transactions are still mostly not automated



Six fundamental stages of the buying process:


Need identification


Product brokering


Merchant brokering


Negotiation


Purchase and delivery


Product service and evaluation



ICT619

30

Intelligent agents in E
-
Commerce
(cont’d)


In the need
-
identification stage, agents can help in
purchases that are repetitive or predictable



Continuously running agents can monitor a set of
sensors or data streams and take actions when certain
pre
-
specified conditions apply



Agents can use rule
-
based systems or data mining
techniques to discover patterns in customer behaviour
to help customers find products

ICT619

31

Intelligent agents in E
-
commerce
(cont.)



In the merchant brokering stage, on
-
line
shopping agents can look up prices for a
chosen product for a number of merchants



Many business
-
to
-
business transactions are
canvassed



In a web auction, customers are required to
manage their own negotiation strategies


Intelligent agents can help with this

ICT619

32

Examples of on
-
line shopping
framework with agent mediation


PERSONA
Logic

Firefly


Bargain
Finder

Auction

Bot

Jango

Auction
Bot

T@T

Need
identification

Product
brokering

*

*

*

*


Merchant
brokering

*

*

*

Negotiation

*

*

*

Payment &
delivery

Service &
Evaluation

ICT619

33

Examples of on
-
line shopping
framework with agent mediation


ICT619

34

Examples of on
-
line shopping
framework with agent mediation


ICT619

35

Examples of on
-
line shopping framework
with agent mediation (cont’d)


Software agents are helping buyers and sellers cope
with information overload and expedite the online
buying process



Agents are creating new markets (eg, low
-
cost
consumer goods) and reducing transaction costs



Use of agents in e
-
commerce still at an early stage



Visit
http://agents.umbc.edu/Applications_and_Software/Ap
plications/Electronic_Commerce/index.shtml


for more

ICT619

36

Intelligent agent design
-

state
-
of
-
the
-
art and future


Few agents are available with all the desired
characteristics


Agent technology still in experimental stage



Autonomy and mobility already achievable




Example: Java applets which execute independently
across networks



But autonomy limited so far in practical use due to the
agent
-
unfriendliness of the current web technology


ICT619

37

Intelligent agent design
-

state
-
of
-
the
-
art and future (cont’d)


A major limiting factor is lack of ontologies
essential for effective communication



Building and maintaining ontologies remains a
major challenge



Some of the proposed capabilities to be
developed in future intelligent agents include:


Learning as well as reasoning, which are
characteristics of machine intelligence


Interacting with the external environment through
sensors

ICT619

38

REFERENCES


Chin, D.,

Intelligent Interfaces as Agents. In
Intelligent User
Interfaces
, J. Sullivan and S. Tyler(eds), ACM Press, New
York, 1991.


Hendler, J.,
Making Sense out of Agents
, IEEE Intelligent
Systems, March/April 1999, pp.32
-
37.


Hendler, J.,
Is There an intelligent Agent in Your Future?
http//www.nature.com/nature/webmatters/agents/agents.html


Maes, P.,
Agents that Reduce Work and Information Overload
,
Communications of the ACM, Volume 37 , Issue 7 (July
1994),
pp.

30
-
40.


Maes, P.,
Agents that Buy and Sell
, Communications of the
ACM, Volume 42 , Issue 3 (March 1999),
pp.

81
-
91.


Sheth, B. and Maes, P. Evolving Agents for Personalized
Information Filtering. In Proceedings of the Ninth Conf. on
Artificial Intelligence for Applications. IEEE Computer Society
Press, 1993


UMBC Agent News
-

http://agents.umbc.edu/agentnews/current/


http://www.agentland.com/