THE DATA ADMINISTRATION NEWSLETTER – TDAN.com
ROBERT S. SEINER – PUBLISHER

Published in TDAN.com January 2003

Business Rule Analysts find a business rule classification scheme to be a useful tool in developing a set of business rules that is complete for the given business domain and consistent with one another. In this series, we have been examining a business rule classification scheme that uses English grammar as a metaphor for an understanding the statements made by the domain area’s Subject Matter Experts (SME).

This Business Rule Grammar consists of two primary components: Business Concepts and Business Rules. A Business Concept is something that the business must know to conduct business. They correspond to English language nouns or qualified nouns, typically expressed through the use of adjectives. Business Rules are expressed as statements about the business concepts, usually in the form of individual sentences. The syntax of the sentence is often an indication of what type of business rule the SME is communicating. Consequently, a solid understanding of English grammar can be an important skill for a Business Rule Analyst.

The two primary types of business concepts are thingees and characteristics. A thingee is anything that the enterprise must know about to successfully conduct business. A characteristic is something that the enterprise must know about one of those thingees. Each of these business concept categories is comprised of a hierarchy of more discrete business concept categories (Table 1).

The first two articles in this series focus on Thingee business concepts (Business Rules Grammar, Part 1, The Data Administration Newsletter) and the business rules that primarily involve those thingees (Business Rule Grammar, Part 2, The Data Administration Newsletter). This article delves into Characteristic business concepts and their associated business rules.

Characteristic Business Concepts

In most cases, a characteristic specifies the properties about a piece of information about a specific thingee. In this situation, the characteristic is a Knowledge Property because it can take on a value. It is a holder of data. For examples, Name, Gender, Birth Date, Age, Social Security Number and Driver’s License are all Knowledge Properties that hold a value appropriate for a specific person. In fact, the more appropriate names for these knowledge property business concepts are Person’s Name, Person’s Gender, Person’s Birth Date, etc. The use of the possessive in a SME’s statement can be an indicator to the business rule analyst that a knowledge property has been discovered.

Name, Gender, Birth Date and Age can apply to other thingees besides Person. All of these can equally apply to my cat. Pet’s Name, Pet’s Gender, Pet’s Birth Date and Pet’s Age are all knowledge properties. But some business concepts that behave like a characteristic, but can’t hold a value because they exist independent of any Thingee. However, these types of characteristics may serve as the basis for defining the properties of other characteristics. The Name, Gender, Birth Date and Age business concepts can be thought of as templates that can be used as the basis of defining knowledge properties are similar to one another.

For example, almost all knowledge properties that are names have a similar definition and similar domain ( for example, “Text that is used to reference the associated object alphanumeric” and maximum length 50). By classifying a thingee’s Name knowledge property to the Name characteristic classifier, the Name knowledge property inherits the properties specified for the classifier. Any property can be overridden at the knowledge property level. For example, I may decide that Person’s Name can only contain alphabetic characters and Product Name is unique across all products. When a knowledge property has been classified, its classification can serve as a keyword in a search. You can ask to view all characteristics classified as a Name, even if the business concepts do not have the term “Name” in their name.

Characteristic classifiers can be considered by Data Management as the set of business rules that are candidates to be supported by Classwords in their attribute and column naming standards.

Characteristic-Related Business Rules

Some business rule categories are primarily oriented toward stating properties, associations, expected behavior or constraints associated with characteristics. These business rules specify how characteristics are classified, associated with their thingees, grouped together into a collections and describe their domain and derivation algorithms (Figure 1).

Characteristic Classification

As already discussed, a Characteristic Classifier specifies properties that are inherited by all the characteristics that it classifies. The Characteristic Classification business rule associates a specific Characteristic with its Characteristic Classifier. Every Knowledge Property should have a classifier identified for it. In addition, a Characteristic Classifier may be classified by another classifier, forming a classification hierarchy, through a series of Characteristic Classification business rules.

For example, consider the situation in which four characteristic classifiers (Memo, Note, Description and Comment) and three knowledge properties (Product Description, Product Notes and Event Comment) exist (Table 2). As the definition of each characteristic, there are nuances between each of the characteristic classifiers that provide guidance in determining which classifier should be associated with a knowledge property.

The Characteristic Classification business rule is used to classify a characteristic. These classifiers are linked together into a hierarchy through successive Characteristic Classification business rules (Figure 2).

A Business Rules Management (BRM) product should understand the inheritance between the parent and child characteristics that participate in a Characteristic Classification business rule. When specifying the details of the child characteristic, any properties defined for the classifying characteristic should automatically be made available for use by the classified characteristic, including any Domain and Derivation business rules. You should have the ability to override these properties and business rules to meet the specific needs of the child characteristic.

Thingee Knowledge Property

A Thingee Knowledge Property business rule associates a specific Knowledge Property to the Thingee it describes.

Additional properties should be defined as part of this business rule that specifies the interaction between the knowledge property and its thingee. Is the knowledge property optional or mandatory? In other words, can the Thingee exist if a value for this knowledge property is not established? Can the knowledge property assume more that one value at the same time (is it multi-valued?)? If so, is there any additional information that must be known about each value, such as effective dates, the date the value was set or who set the value? If so, there may exist some additional business concepts lurking behind this knowledge property. Can the value be changed once it is set?

The answers to these questions will guide in determining the support for the knowledge property within the information resource. In most cases, the Thingee Knowledge Property business rule will be supported as an attribute of a data model entity or a class in an object class model.

Knowledge Property Configuration

Often, the SMEs will identify a set of knowledge properties that should be treated as a group with a specific configuration. For example, the knowledge property “US phone number” consists of the following knowledge properties: “Country Code”, “Area Code”, “Prefix” and “Line Number”. To be meaningful to the business, the components of US Phone Number must always be presented and manipulated with this specific configuration. A knowledge property collection may also be specified in support of a specific business process. In this situation, the collection may be used in conjunction with a form or report in which each knowledge property must appear in a specific position.

The Knowledge Property Configuration business rule is used to specify a set of Knowledge Properties that the business treats as a unit. Properties that can be specified include the position of each a knowledge property within the collection and the rules governing the individual knowledge properties within the collection. For example, the last 4 digits of the US Zip Code are optional.

In many cases, a Knowledge Property Configuration business rule will be supported by a view into the data or object class model.

Characteristic Domain

Every characteristic has a domain that constrains the set of values that it can assume. A domain specifies the data types of the knowledge property’s values, such as numeric, character, boolean, date, image or sound. Additional properties, relevant to the data type, can also be provided. For example, a numeric range may be specified for which the value must fall into.

The Characteristic Domain business rule specifies the set of values that a Knowledge Property can assume and the conditions under which it can assume those values. Additional properties that can be specified as part of this business rule are the preferred formats. For example, the SMEs state that the US phone number is to be formatted with the area code is surrounded by parentheses and a dash appearing between the prefix and line number knowledge properties. They may also state that, if the last 4 digits of the US Zip Code are available, then the Zip Code’s format is 99999-9999.

A Characteristic Domain business rule can be specified for a Characteristic Classifier. This domain is then inherited by all characteristics associated with that classifier.

Derivation

The value of some characteristics can be derived. In this situation, a Derivation business rule is stated expressed in terms of an algorithm can be used to determine the value. The algorithm may use other characteristics as input, as well as mathematical, date and string manipulation functions.

As with the Characteristic Domain business rule, a derivation business rule can be specified for a Characteristic Classifier.

The business rule grammar presented in this series can be used in analyzing the statements made by the SMEs to ensure that each business rule is clear, concise and unambiguous. Additionally, the business rule grammar can be used in discovering and expressing a set of business rules that is complete, for the intended purpose, and in ensuring that the business rules are not in conflict with one another.

Business rules exist to guide and, when possible, control the behavior of the enterprise Business rules are communicated throughout the organization through its policy manuals. Each release of the enterprise’s policy manuals reflect its business rule set at a specific point in time. Likewise, the enterprise’s procedure manuals identify its business processes and describe how it expects to behave when specific business events occur.

If Business processes prescribe expected behavior and business rules are intended to guide behavior, then a symbiotic relationship exists between business rules and business processes. That relationship will be the subject of the final article in this series.