Required reading
	---------------------------------------------------------
	Textbook	Laudon & Laudon 1995 Ch. 14. pp. 506-530 Ch. 15. pp. 546-553, 563-564, 568-572
	---------------------------------------------------------

Overview

Chapter 14 discusses 'knowledge work' and office automation technology.

From Chapter 15 you should gain an introduction to artificial intelligence, expert systems and neural networks. There have been a number of people at the University doing research in these areas for several years, including me. I have developed systems for Capital Gains Tax and financial planning. I've also looked at interface issues associated with expert systems. Other people have developed systems for agriculture, drug dosage and building construction.

Review questions

REVIEW QUESTION 9-1	Textbook	Laudon & Laudon 1995 pp. 543-545, Groupware case study, Questions 1 to 4.
REVIEW QUESTION 9-2	Textbook	Laudon & Laudon 1995 p. 578, Discussion questions 1 to 3.

Quiz

Mark the correct answer.

QUESTION 1

Which of the following is NOT a role of the office?
(a) coordinates work of business professionals
(b) develops long-range plans
(c) links diverse parts of the business
(d) spans boundaries between the business and its external environment
(e) couples firm to its clients and suppliers

QUESTION 2

An exact copy of a document may be transmitted over telephone lines using a:
(a) scanner
(b) modem
(c) facsimile
(d) plotter
(e) multiplexer.

QUESTION 3

Which of the following groups do NOT normally work in the office?
(a) sales
(b) professional
(c) managerial
(d) production
(e) clerical

QUESTION 4

The normal printer used with a desktop publishing package is what kind of printer?
(a) dot matrix
(b) character
(c) laser
(d) plotter
(e) line

Put a tick in the appropriate box.	True	False
QUESTION 5
If you look carefully, you will find that offices are predominantly oriented to secretaries and clerical workers.
QUESTION 6
Databases track the calendars of related individuals in an office and make those calendars available to the central receptionist.
QUESTION 7
Groupware attempts to expedite all the functions of a typical work group.
QUESTION 8
Technological and technical changes have made employees much less important in knowledge systems.
Write the most appropriate word on the line.
QUESTION 9
High-quality output, including graphics and layout,is part of ___________________.

Answers to quiz

Question 1.	(b)
Question 2.	(c)
Question 3.	(d)
Question 4.	(c)
Question 5.	F
Question 6.	F
Question 7.	T
Question 8.	F
Question 9.	desktop publishing

Answer pointers to Review questions

Taken from Simmons (1995).

REVIEW QUESTION 9-2

1. Since expert systems cannot learn, it is unlikely that they will ever replace upper management. Expert systems must be reprogrammed whenever knowledge changes in a field. Because top managers face a rapidly changing business environment, expert systems could not perform their function without prohibitive reprogramming costs. Furthermore, much expertise possessed by human managers cannot be organized in the IF THEN format required by expert systems. This effectively limits expert systems to classification problems and severely restricts their ability to solve open-ended problems, the type managers most often face. It is more likely that human managers could be replaced by neural networks which are more flexible and can be adapted to a wider variety of problems as well as to a changing environment.

2. An expert system might be used to enable the college financial aid department to operate more efficiently. An expert system is appropriate because the department's major task is classification of students into categories; either eligible or ineligible for various grants and loans. A knowledge engineer would interview experts, the financial aid officers, to develop a set of rules which model the officer's decision-making process. These rules would then be encoded in an AI shell enabling clerical employees to enter into the system relevant student financial data. The expert system would generate a recommended category for each student. Finally, the financial aid officer would accept or reject the category based on the computer's analysis. This would reduce the amount of time a financial aid officer spends gathering and analyzing student information.

3. Key decisions could be left to artificially intelligent devices if the user could be certain the device would always make the "correct" decision. However, AI devices are only as accurate as their programming. Since programming is created by imperfect humans, it is likely to have mistakes. The only way to detect these errors is to reserve key decision making for humans. This allows users to examine the computer's recommendation and analysis for errors before the ultimate decision is made. If key decisions are left to AI devices, errors will not be discovered until after the "wrong" decision has been made.

   Another consideration is that many high-level decisions are too open ended and complex to be captured by expert systems and will always require human expertise. However, expert systems could assist human decision makers by performing some of the routine analysis and tedious work while the human thinks or by coming up with some of the alternatives for the human decision maker to evaluate.