Classification by Intelligence level and power
Lee et al. classify Web agents into four levels, according to their
intelligence and power.
1) Level 0 These agents retrieve documents for a user under straight
orders. Popular Web browsers fall into this category.
2) Level 1 These agents provide user initiated searching facility for
finding relevant Web pages.
Internet search agents such as Yahoo, Alta Vista, and
Infoseek are examples.
When the user provides key words, the search engine
matches them against the indexed information.
3) Level 2 These agents maintain user’s profiles. Then they monitor
internet information and notify the users whenever relevant
information is found.
Examples of such agents are WebWatcher and SIFT.
4) Level 3 Agents at this level have learning and educative component
of user profiles to help a user who cannot formalize a query
or specify a target for a search.
DiffAgent and Letzia are examples of such agents.
Classification of application area
According to IBM’s white paper, there are nine major application areas
that elate to internet agents.
1. Agents that assist in workflow and administrative management
2. Agents that collaborate with other agents and people
3. Agents that support electronic commerce
4. Agents that support desktop applications
5. Agents that assist in information access and management
6. Agents that process mail and messages
7. Agents that control and manage the network access
8. Agents that manage systems and networks
9. Agents that create user interfaces
Gilbert and Janca indicates that many agents are being used in each
category, be in mainstream use. They also provide many examples of
agents in each of the nine categories.
King classifies agents into interface, tutors, scheduling assistants,
search agents, report agents, presentation agents, navigation agents,
and role playing agents.
Internet based software agents
Following are the main internet based software agents.
1. E-mail agents
2. Web browsing assisting agents
3. Frequently asked questions agents (FAQ)
4. Intelligent search agents
5. Internet softbot
Managerial Issues
The following are some representative managerial issues.
1) Cost Justification 5) Other ethical issues
2) Security a) Agent learning
3) Privacy b) Agent accuracy
4) Industrial intelligence and ethics c) Heightened expectations
5) Other ethical issues d) System acceptance
e) System technology
f) Agent development toolkits
g) Strategic information systems
Cost Justification
With technology rapidly changing and with intelligent agents in their
developmental infancy, it ma be hard to justify spending lots of money
on something that may be obsolete tomorrow.
Security
Agents are a technology with many unknown ramifications.
With the great concerns about the security the security of systems,
does it make sense that a company would knowingly send out agents
that could come back laden with a virus or hiding a Trojan horse?
Privacy
There are some cases in which agents have intruded on people’s
privacy.
Example _ Cookies
Industrial intelligence and ethics
Legitimate industrial intelligence gathering is usually expensive and
time consuming.
Other ethical issues
Agents represent a significant new way of interacting with the world.
Just as a unique etiquette has evolved on the internet, there will need
to be new definitions of acceptable and unacceptable uses for agents.
Agent learning
The theory behind agents is that the more you use them, the more they
learn and, therefore, the more effective they are.
Agent accuracy
Along with agent learning, comes agent accuracy.
Assuming that the agent develops the ability to learn, the next issue
facing managers is the accuracy of the data returned by the agent.
Heightened expectations
With any new technology or product come higher expectations. This is
especially true when it comes to intelligent agents.
System acceptance
The introduction of intelligent agents into an existing system can
sometimes create problems.
Systems have different architectures and operating systems.
System technology
As intelligent agents become more powerful, the systems required to
run them also must be more powerful.
Agent development toolkits
To implement agent applications, one can use specially toolkits.
Strategic information systems
Management has a great need to survey the environment for
opportunities and treats to the long term survival and success of the
company.
Intelligent agents are an excellent long term strategic asset a company
may develop in gathering data from a variety of sources around the
world.
EDP Electronic Data Processing
UIMS User Interface Management System
GUI Graphical User Interface
IFPS Interactive Financial Planning System
DFPM Educom’s Financial Planning Model
ES Expert System
ANN Artificial Neural Networks
VRML Virtual reality Markup Language
GPS Global Positioning System
GIS Global Information System
OLE Object Linking and Embedding
NGT Nominal Group Technique
SANN Simulated Artificial Neural Networks
Lee et al. classify Web agents into four levels, according to their
intelligence and power.
1) Level 0 These agents retrieve documents for a user under straight
orders. Popular Web browsers fall into this category.
2) Level 1 These agents provide user initiated searching facility for
finding relevant Web pages.
Internet search agents such as Yahoo, Alta Vista, and
Infoseek are examples.
When the user provides key words, the search engine
matches them against the indexed information.
3) Level 2 These agents maintain user’s profiles. Then they monitor
internet information and notify the users whenever relevant
information is found.
Examples of such agents are WebWatcher and SIFT.
4) Level 3 Agents at this level have learning and educative component
of user profiles to help a user who cannot formalize a query
or specify a target for a search.
DiffAgent and Letzia are examples of such agents.
Classification of application area
According to IBM’s white paper, there are nine major application areas
that elate to internet agents.
1. Agents that assist in workflow and administrative management
2. Agents that collaborate with other agents and people
3. Agents that support electronic commerce
4. Agents that support desktop applications
5. Agents that assist in information access and management
6. Agents that process mail and messages
7. Agents that control and manage the network access
8. Agents that manage systems and networks
9. Agents that create user interfaces
Gilbert and Janca indicates that many agents are being used in each
category, be in mainstream use. They also provide many examples of
agents in each of the nine categories.
King classifies agents into interface, tutors, scheduling assistants,
search agents, report agents, presentation agents, navigation agents,
and role playing agents.
Internet based software agents
Following are the main internet based software agents.
1. E-mail agents
2. Web browsing assisting agents
3. Frequently asked questions agents (FAQ)
4. Intelligent search agents
5. Internet softbot
Managerial Issues
The following are some representative managerial issues.
1) Cost Justification 5) Other ethical issues
2) Security a) Agent learning
3) Privacy b) Agent accuracy
4) Industrial intelligence and ethics c) Heightened expectations
5) Other ethical issues d) System acceptance
e) System technology
f) Agent development toolkits
g) Strategic information systems
Cost Justification
With technology rapidly changing and with intelligent agents in their
developmental infancy, it ma be hard to justify spending lots of money
on something that may be obsolete tomorrow.
Security
Agents are a technology with many unknown ramifications.
With the great concerns about the security the security of systems,
does it make sense that a company would knowingly send out agents
that could come back laden with a virus or hiding a Trojan horse?
Privacy
There are some cases in which agents have intruded on people’s
privacy.
Example _ Cookies
Industrial intelligence and ethics
Legitimate industrial intelligence gathering is usually expensive and
time consuming.
Other ethical issues
Agents represent a significant new way of interacting with the world.
Just as a unique etiquette has evolved on the internet, there will need
to be new definitions of acceptable and unacceptable uses for agents.
Agent learning
The theory behind agents is that the more you use them, the more they
learn and, therefore, the more effective they are.
Agent accuracy
Along with agent learning, comes agent accuracy.
Assuming that the agent develops the ability to learn, the next issue
facing managers is the accuracy of the data returned by the agent.
Heightened expectations
With any new technology or product come higher expectations. This is
especially true when it comes to intelligent agents.
System acceptance
The introduction of intelligent agents into an existing system can
sometimes create problems.
Systems have different architectures and operating systems.
System technology
As intelligent agents become more powerful, the systems required to
run them also must be more powerful.
Agent development toolkits
To implement agent applications, one can use specially toolkits.
Strategic information systems
Management has a great need to survey the environment for
opportunities and treats to the long term survival and success of the
company.
Intelligent agents are an excellent long term strategic asset a company
may develop in gathering data from a variety of sources around the
world.
EDP Electronic Data Processing
UIMS User Interface Management System
GUI Graphical User Interface
IFPS Interactive Financial Planning System
DFPM Educom’s Financial Planning Model
ES Expert System
ANN Artificial Neural Networks
VRML Virtual reality Markup Language
GPS Global Positioning System
GIS Global Information System
OLE Object Linking and Embedding
NGT Nominal Group Technique
SANN Simulated Artificial Neural Networks
No comments:
Post a Comment