Chapter 7. An Evolving Database

Every day we use formal and informal rules to control our actions. We know that a hot flame will burn our skin and will hurt. Therefore, we know not to touch the flame. Automobile drivers know that if they drive too fast, they may get a speeding ticket.

Businesses also utilize rules to control the actions they take. Business rules encapsulate the logic and procedures used by businesses to handle certain situations. Such rules can exist formally in policy manuals or informally as the experiences of employees. An example of a business rule involving a small widget retailer would be, "If the quantity of widgets falls below 200, then order 500 more widgets."

Rule-based systems seek to capitalize on the knowledge of experts by defining a set of rules used to represent a particular area. In the case of the widget retailer, the experts would be store managers, buyers, or stock clerks. These employees all understand the particular facets of their jobs. They would therefore be the most knowledgeable when it comes to defining how their jobs should be done.

The set of rules used by these employees can overlap and may vary depending on slight differences. For instance, the rule about ordering more widgets may change if it is November because widgets do not sell well in that month.

Every rule is associated with actions, and once a rule is defined, it is applied to a set of inputs. The inputs are such factors as the quantity of widgets or the current month. The end result or action applied depends on how the rules are applied to the inputs.

In the financial industry, rule-based systems have been very helpful in decision-making about stock performance or creditworthiness. These systems can be used either to enhance a human's deductive capabilities or, in some cases, to replace them. In today's 24/7 world of instant accessibility, it is sometimes vital for companies to be able to make decisions involving customers as quickly as possible. Rule-based systems allow companies to "bottle" their employees' decision-making capabilities so they will be available for quick and efficient use.

This chapter will examine a rule-based system used by a fictional financial lender named Zoom Lending. The ASP.NET application is used by employees to create and manage borrower surveys. The program contains a simplified rule engine used to determine when new survey questions should be added as new fields in the database.

Rule-based systems can involve just about any situation in which an automated response requires that a decision be made. The rules are dynamic, and different ones come into play depending on the input. The application in this chapter is a simple rule-based system in which the expert is the database administrator (DBA). Since decisions involve the addition of fields to the database, the DBA is the person most knowledgeable of how those fields should be defined.

The chapter also contains profile boxes featuring two Microsoft products that utilize rule-based systems. The Microsoft Outlook rules wizard, part of the Microsoft Outlook product family, can automatically manage messages using user-defined rules. BizTalk Server 2004, now in its third major release, features a way to create and modify business rules through a graphical interface that can be used even by those who are not developers.