Page 400 -

399
Q9-7 What Is the Role of Knowledge Management Systems?

To create the system of rules, the expert system development team interviews human experts
in the domain of interest. The rules in Figure 9-25 would have been obtained by interviewing
cardiologists who are known to be particularly adept at diagnosing cardiac disease. Such a system
encodes the knowledge of those highly skilled experts and makes it available to less-skilled or
less-knowledgeable professionals.
Many expert systems were created in the late 1980s and early 1990s, but only a few have
enjoyed success. They suffer from three major disadvantages. First, they are difficult and expen-
sive to develop. They require many labor hours from both experts in the domain under study and
designers of expert systems. This expense is compounded by the high opportunity cost of tying
up domain experts. Such experts are normally some of the most sought-after employees in an
organization.
Second, expert systems are difficult to maintain. Because of the nature of rule-based sys-
tems, the introduction of a new rule in the middle of hundreds of others can have unexpected
consequences. A small change can cause very different outcomes. Unfortunately, such side effects
cannot be predicted or eliminated. They are the nature of complex rule-based systems.
Finally, expert systems were unable to live up to the high expectations set by their name.
Initially, proponents of expert systems hoped to be able to duplicate the performance of highly
trained experts, like doctors. It turned out, however, that no expert system has the same diagnostic
ability as knowledgeable, skilled, and experienced doctors. Even when expert systems were developed
that came close in ability, changes in medical technology required constant changing of the expert
system, and the problems caused by unexpected consequences made such changes very expensive.
The few expert systems that have been successful have addressed more restricted problems
than duplicating a doctor’s diagnostic ability. They address problems such as checking for harm-
ful prescription drug interactions and configuring products to meet customer specifications.
These systems require many fewer rules and are therefore more manageable to maintain.
However, unless expert systems technology gets a boost from massively parallel computing (think
MapReduce and Hadoop), their problems will cause them to fade from use.

What Are Content Management Systems?

Another form of knowledge management concerns knowledge that is encoded in documents.
Content management systems (CMS) are information systems that support the management
and delivery of documents including reports, Web pages, and other expressions of employee
knowledge.
Typical users of content management systems are companies that sell complicated products
and want to share their knowledge of those products with employees and customers. Someone
at Toyota, for example, knows how to change the timing belt on the four-cylinder 2015 Toyota
Camry. Toyota wants to share that knowledge with car owners, mechanics, and Toyota employees.

What Are the Challenges of Content Management?

Content management systems face serious challenges. First, most content databases are huge;
some have thousands of individual documents, pages, and graphics. Second, CMS content is
dynamic. Imagine the frequency of Web page changes at Apple or Google or Amazon.com that
must occur each day!
Another complication for content management systems is that documents do not exist in
isolation from each other. Documents refer to one another, and when one changes, others must
change as well. To manage these connections, content management systems must maintain link-
ages among documents so that content dependencies are known and used to maintain document
consistency.
A fourth complication is that document contents are perishable. Documents become obsolete
and need to be altered, removed, or replaced. Consider, for example, what happens when a new
product is announced. Figure 9-26 shows the main page for Microsoft.com less than 2 hours