BI510 Business Rules ARE MetaData

1
BI510Business Rules are Meta-Data

• Alan Perkins
• EVP, Client Services
• Visible Systems
• alan.perkins_at_visible.com

2
Business Rules Meta-Data

• Benefits of Managing Business Rules as Meta-Data
• Techniques for Defining, Documenting, Modeling,
  and Implementing Business Rules
• The Business Rule Life Cycle

3
Business Rule Definition

• A Business Rule ... defines or constrains some
  aspect of the business. It is intended to assert
  business structure or to control or influence the
  behavior of the business.
• -- IBM GUIDE Business Rules Project

These ARE Meta-Data

• Business Rules may be
• Definitions of Business Terms
• Data integrity constraints
• Mathematical and functional derivations
• Logical inferences
• Processing sequences
• Relationships among facts

4
Uses for Business Rules

• Capture/Implement Precise Business Logic
• Provide Basis for Expert Systems
• Refine Architecture Models and Software
  Component Designs
• Facilitate Change Management
• Manage XML Applications
• B2B Communications
• Corporate Portals
• Application Integration

5
Implementing Business Rules

• As Procedure (Manuals, SOPs, etc.)
• Flexible
• But not rigorous
• As Process (Subroutines, object behavior, code)
• Rigorous
• But inflexible
• As Meta-Data (Tables, triggers, and objects)
• Rigorous
• And Flexible

6
Benefits of Managing Business Rulesas Meta-Data

• Allows Maximum Flexibility
• Reduces System Maintenance
• Simplifies Design and Implementation
• Rules can change without affecting implementation
  
• Ensures that systems fully support business needs
• MIS personnel don't need to learn the intricacies
  of the business

7
Business Rules Meta-Data

• Benefits of Managing Business Rules as Meta-Data
• Techniques for Defining, Documenting, Modeling,
  and Implementing Business Rules
• The Business Rule Life Cycle

8
Business Rule Rules
Effective Business Rules have the following
characteristics

• Explicit expression The statement of business
  rules needs an explicit expression, either
  graphically or as a formal (logic-based)
  language.
• Declarative nature A business rule is
  declarative, not procedural. It describes a
  desirable, possible state that is either required
  or prohibited.
• Coherent representation A single, coherent model
  for all the kinds of business rules is desirable.

9
Business Rules Reasons to Model
Ready

• Encourages rigorous and complete analysis.
• Reduces complexity.
• Enhances communication.
• Assists in eliminating the Ready - Fire - Aim
  mentality.

Fire
Aim!
OOPS
You should know your targetbefore you pull the
trigger.
10
Enterprise Business ModelBusiness Rule Repository

• Incorporates best practices.
• Meets the unique needs of the enterprise
• Eliminates redundant analysis.
• Promotes data sharing.
• Encourages reuse -- not reinvention.
• Allows effective component development AND
  management

11
Business Models Provide Framework for Component
Design Development
StrategicandPerformancePlans
Strategic Business Rules
FunctionModel
InformationModel
Business Profiles
Tactical Business Rules
LogicalComponentModel
LogicalDataModel
Site Profiles
Operational Business Rules
PhysicalComponentDesign
PhysicalData BaseDesign
Site
Component Code
Data Base
12
Enterprise Architecture Frameworkeverything you
need to know...
Data
Process
Infrastructure
Environment
Business Plan
Business Events
Enterprise Context
Internal/External Assessment
Business Plan
Business Requirements Model
Information Model
Activity Model
Environmental Components
Enterprise
Business Model
Model View Locations
Process Model
Data Model
Model Views
Systems Design
Data Base Designs
Technology Configurations
Infrastructure Designs
Application Designs
Sites Platforms
Information System
Assignments Authorizations
Application Systems
Data Bases
13
The Only Enterprise Architecture Tool Set
Visible Advantage
Logical Enterprise (Data Warehouse)Model
14
One Logical Model -- Many Physical Designs
Model Views
Logical ModelIncludesBusinessRequirements
Design Views
Each can represent a different technology Each
can be differently de-normalized May be new or
legacy designs
15
Business Rules in Business Modelsthe ultimate
meta-data

• Documented using
• Business Statements
• Data Entities and Attributes
• Entity Associations
• Domain Values
• Derivations
• Rule Expressions

16
Development of Business Statements
Enterprise Mission
Critical Success Factors
Events
Critical Decision Set
Strategies
Goals
Value-Based Process
Objectives
Critical Assumption Set
Performance Measure
Information Needs
Functions
Business Rules
17
Business Statements Guidelines

• Write clear, concise statements in business
  language
• State business, not systems requirements
• Identify each statement clearly
• Resolve conflicts and overlaps
• Fill any gaps

18
Business Statement Hierarchy
19
Business Rules in Business Modelsthe ultimate
meta-data

• Documented using
• Business Statements
• Data Entities and Attributes
• Entity Associations
• Domain Values
• Derivations
• Rule Expressions

20
Entity Definition

• Some "thing" which is of importance to the
  business
• Conventions
• Rectangular box
• Named in the singular
• Capital letters
• Examples

21
Attribute Definition

• A "piece" of information which describes an
  entity
• Conventions
• Named in the singular
• Written in lower case
• Examples

22
Attribute Primary Keys

• Uniquely identifies the occurrence within an
  entity
• Conventions
• identifies attribute as a key attribute
• Primary key is underlined
• Examples

23
Attribute Foreign Keys

• A reference to an occurrence in another entity.
  It is the primary key of that other entity
• Convention
• identifies attribute as a key attribute
• Example

24
Attribute Non-Keys

• Additional information about an entity
• May be defined as derived, unique or optional
• Convention
• Lowercase with no identifying characteristic
• Examples

25
Business Rules in Business Modelsthe ultimate
meta-data

• Documented using
• Business Statements
• Data Entities and Attributes
• Entity Associations
• Domain Values
• Derivations
• Rule Expressions

26
Association Definition

• Indicates relationship between two entities
• Conventions
• A line connecting two entities
• Only one association exists between any two
  entities
• (There may be multiple reasons for the
  association)

27
Association Degree

• Identifies whether an entity may be related to
  one or more occurrences of other entities
• Conventions

One or more
Only one
28
Association Nature

• Defines whether an entity may exist on its own or
  must be related to another entity
• Conventions

Optional
Mandatory
Optional becoming
mandatory
29
Combinations of Degree Nature
An Employee ....
must have one
A
must have at least one
B
may have zero or one (or may be a...)
C
EMPLOYEE
D
may have zero, one or many
E
will eventually have one
F
will eventually have at least one
30
Association Examples

• Each employee must have one and only one job. A
  job must be related to one or more employees.
• An employee will eventually be associated with a
  paycheck. Each paycheck relates to only one
  employee.

31
Multiple Associations
32
Multiple Associations Resolution
33
Alternative Solution ...
34
Business Rules in Business Modelsthe ultimate
meta-data

• Documented using
• Business Statements
• Data Entities and Attributes
• Entity Associations
• Domain Values
• Derivations
• Rule Expressions

35
Types of Domains

• Entity
• Meta-data values
• Attribute
• Data type values

36
Entity Nature

• Static
• Limited, relatively small number of occurrences,
  with known values
• Dynamic
• Unlimited number of occurrences, with unknown
  values

37
Populating a Static Entity
38
Attribute Domain Concepts

• Each Attribute must eventually be assigned a
  Domain
• Should not be constrained by technology
  limitations
• Domains may overlap

Attribute B
Attribute A
Domain
Domain
39
Domain Concepts

• The domain of an attribute define the values that
  an attribute may have
• The actual values of an attribute may be a subset
  of the actual domain

Attribute B
Attribute A
Actual values
Actual values
Domain
Domain
40
Domain Concepts

• A domain can be assigned a standard length and
  precision for all attributes that use that domain
  - Global Dimension
• Local Dimension - allows length to be specified
  at the Attribute level. Example Character
  domain, length 20

41
Composition of Domains

• Range
• Discrete
• Complex

Composed of an infinite set of values, may be
bounded at either end Composed of a finite
number of known values Consists of more than one
group of discrete and/or range sets
42
Domain Examples

• Discrete
• January, February, March, April, May, June,
  July, August, September, October, November,
  December
• ABCDEFGHIJKLMNOPQRSTUVWXYZ
• 50 state abbreviations
• High, Medium, Low (3-level categorization)
• Century, Year, Quarter, Month, Week, Day, Hour,
  Minute, Second

43
Domain Examples

• Range
• Positive Integers
• U.S. Currency Amount (to nearest )
• Names
• 50-500 lbs (person's weight)

44
Data Types

• Logical, independent of any specific technology
• Provides a format for a Domain
• Shared by many Domains
• Standard Data Types include
• Numerical, Character, Boolean
• Date, Time
• Picture, Video, Sound

45
Business Rules in Business Modelsthe ultimate
meta-data

• Documented using
• Business Statements
• Data Entities and Attributes
• Entity Associations
• Domain Values
• Derivations
• Rule Expressions

46
Attribute Derivation

• An attribute whose value is determined by the
  existence or values of one or more other
  attributes.
• For each derived attribute a derivation formula
  must be specified. The derivation formula
  identifies the relevant attributes and the
  calculation necessary to compute a value for the
  derived attribute.

47
Functions

• Predefined algorithm on a variable set of
  arguments
• Examples
• (Numerical) COUNT (ENTITY)
• Counts the occurrences of an entity, returned as
  a numerical value
• (Numerical) SUM (Numerical)
• Sums all the occurrences of an attribute

48
Unary Operators

• Unary operators have only one operator
• Examples
• (Numerical) signing functions
• -(Numerical)

49
Binary Operators

• Binary operators always have two operators
• Values returned may be
• Boolean e.g. (Numerical) !(Numerical)
• Numerical (Numerical) (Numerical)
• Character (Character) (Character)
• Date (Date) - (Numerical)

50
Derived Attribute Example
51
Derived Attribute Example
52
Business Rules in Business Modelsthe ultimate
meta-data

• Documented using
• Business Statements
• Data Entities and Attributes
• Entity Associations
• Domain Values
• Derivations
• Rule Expressions

53
Requirements for Rule Modeling

• A Stable Data Model
• Populated Static Entities
• Knowledge of the Business

54
Identifying Potential RulesExamine model for
combinations of data

• Intersecting Entities
• - combinations of two (or more) different
  entities
• Principal Entities
• - combinations of foreign keys from static
  entities
• Role Entities
• - combinations of optional data for an occurrence
  of a principal
• Structure Entities
• - combinations of different occurrences of a
  principal

55
Categories of Rule Logic

• Simple Rules
• Used to enforce or preclude the existence of
  specific patterns of relational logic
• Complex Rules
• Used to control the amplitude of occurrences of
  a specific pattern of business logic

56
Simple Rules

• Description
• The simplest control mechanism that can be
  applied to combinations of other data
• Values
• Combinations of values for two attributes may be
  labeled as
• Mandatory (M)
• Not Allowed (N)
• Optional (O)

57
Complex Rules

• Controls combinations more complex than simple
  "Yes/No", "On/Off"
• Used to amplify the degree and/or nature of an
  association
• May encompass nature rules
• Values for minimum and maximum
• - 0
• - 1
• - X (a fixed integer)
• - N (indicates no fixed maximum)

58
Complex Rules Options
Min 0 0 0 0 1 1 1 X
Max 0 1 X N 0 1 X N
Rule Not allowed Optional, may have
one Optional, may have up to X Optional, may
have more than one Not meaningful
(illogical) Mandatory, must only have
one Mandatory, may have up to X Mandatory, must
have X or more
59
Rule Entity

• Usually an INTERSECTING Entity between two static
  Entities. Controls population of these static
  values in another entity
• May be used for either simple or complex rules
• Must have at least a "Read Only" Functional Area
  Authority in each Functional Area that can add or
  change occurrences in the controlled Entity

60
Rule Example for Two Type Entities
PERSON ASSIGNMENT RULE controls PERSON
ORGANIZATION
61
Attribute Combinations Example
62
TYPE STRUCTURE Entity

• A STRUCTURE Entity which captures business rules
  governing the population of another entity
• Used to control the valid combinations of
  categories (types) in one of two patterns
• Population of a Role entity related to a single
  occurrence of Principal entity
• Population of a Structure entity related to two
  occurrences of a Principal entity

63
TYPE STRUCTURE Entity

• Business Rules
• A person who is an employee may not also be a
  contact
• A person who is an employee may not also be a
  relative

Since this rule governs only one occurrence of
the principle, the sequence of Person Type Id's
does not provide an alternate interpretation of
the rule.
64
TYPE STRUCTURE Controlling a STRUCTURE

• Business Rules
• Every employee must be associated with another
  employee
• Every contact must be associated with an employee

• Since the rule governs two occurrences of the
  principal entity, this rule could be read as
• Every contact must be associated with an
  employee, or
• Every employee must be associated with a contact.
• Capture the valid direction of the rule in the
  purpose for the TYPE STRUCTURE entity.

65
Complex TYPE STRUCTURE Rule

• If the TYPE STRUCTURE entity is required to
  control both a ROLE and a STRUCTURE for the same
  TYPE entity, then the occurrences in the TYPE
  STRUCTURE must be categorized

66
REASON Entity

• A PRINCIPAL Entity that contains descriptions of
  the categories of relational logic being captured
  in some other Entity
• Replaces multiple associations between two
  entities
• The rules are normally "directional" indicating
  amplification of the nature, and maybe degree, of
  one association

67
REASON Entity Controlling a STRUCTURE Entity
68
Multiple Use Intersection Rule
69
Alternative to Multiple Use Intersecting Entity
Rule
70
Documenting Rules in the Data Model

• All Rule-based Entities should be linked to the
  management statement that contains the business
  rule being modeled
• All Rule-based Entities should have a Controlling
  Entity link to the Entity(s) that they control
• Rule-based Entities should be designated as
  static, and be populated

71
Business Rules Meta-Data

• Benefits of Managing Business Rules as Meta-Data
• Techniques for Defining, Documenting, Modeling,
  and Implementing Business Rules
• The Business Rule Life Cycle

72
Business Rule Life Cycle
PLANNING REQUIREMENTS
DATA MODELING
OPERATIONAL
STRATEGIC
TACTICAL
FUNCTION MODELING
ACTIVITY
PROCESS
USE CASE
COMPONENT DESIGN
APPLICATION
DATABASE
OBJECT
SYSTEM IMPLENTATION
73
Business Rule Life Cycle Controlled Iteration
of Tasks Techniques
PLANNING REQUIREMENTS
DATA MODELING
OPERATIONAL
Technology Independent
STRATEGIC
TACTICAL
Business Driven Change
FUNCTION MODELING
ACTIVITY
PROCESS
USE CASE
COMPONENT DESIGN
APPLICATION
DATABASE
OBJECT
Technology Dependent
SYSTEM IMPLENTATION
74
Business Rule Life Cycle
PLANNING REQUIREMENTS
PLANNING REQUIREMENTS
StrategicDirection
DATA MODELING
InformationRequirement
OPERATIONAL
Technology Independent
STRATEGIC
TACTICAL
FUNCTION MODELING
InformationUsage
ACTIVITY
PROCESS
USE CASE
COMPONENT DESIGN
Technology/SystemRequirements
Technology Dependent
APPLICATION
DATABASE
OBJECT
PerformanceRequirement
SYSTEM IMPLENTATION
75
Planning Requirements

• The Planning Requirements capability can be
  used to
• Document Strategic and Business Plans
• Document Performance Plans
• Capture reengineered business rules and metrics
  in text form.
• Capture business requirements for information
  systems in plain English
• Document modeling issues with links to the model
  components at issue.
• Document computer system requirements
• Model Use Cases

76
Planning Framework
Mission and Purpose
External Guidelines and Values
Goals
Policies
External and Internal Environment Assessment
Strategies
Measures Metrics
Critical Success Factors Performance Measures
Objectives
77
Strategic Business PlanningThroughout the
Enterprise
StrategicPlan
Corporate Level Sets Enterprise Direction

• Strategic Business Plan
• Mission, Charter and Values
• Strategic Goals and KPIs
• Key Strategies and Priorities
• Tactical or Program Plans
• Key Areas of Focus for each Area
• Strategies Objectives
• Operational Business Plans
• Tasks Measures
• Short term
• Performance Agreements
• KPIs at each Management Level

PerformancePlans
Business Unit Major Enterprise Activities
BusinessPlans
Functional Area Work Groups Sections
PerformanceAgreements
78
Strategic Business PlanningThroughout the
Enterprise
StrategicPlan
Corporate Level Sets Enterprise Direction

• Concerns and Issues
• What opportunities (problems) may enhance
  (impede) effective achievement of business plans?
• Provide feedback, verification, and validation

PerformancePlans
Business Unit Major Enterprise Activities
BusinessPlans
Functional Area Work Groups Sections
PerformanceAgreements
79
Planning Requirements
80
Planning Requirements
81
Plan components (statements) are linked to the
data model and also to the system(s) that
implement the plans.
82
Business Rule Life Cycle
PLANNING REQUIREMENTS
StrategicDirection
InformationRequirement
Technology Independent
InformationUsage
Technology/SystemRequirements
Technology Dependent
PerformanceRequirement
SYSTEM IMPLENTATION
83
Data Modeling Business Requirements
Management Statements
Data Model
84
Review employee skill levels on a quarterly basis
and reallocate personnel immediately to associate
new job skill requirements.
85
Logical Data Model
86
Data models should have textual and graphical
display versions. Group and derived attributes
are necessary
87
Matrices provide the ability to view, edit, and
print links in a straightforward and convenient
manner.
The CRUD matrix (aka Entity/View matrix) is an
effective Data Administration tool. The other
matrix shows the links between statements and
data model components.
88
Business Rule Life Cycle
PLANNING REQUIREMENTS
StrategicDirection
InformationRequirement
Technology Independent
InformationUsage
Technology/SystemRequirements
Technology Dependent
PerformanceRequirement
SYSTEM IMPLENTATION
89
Function/Process Modeling
INFORMATION MODEL Data or Objects
Input / Output / Control
BUSINESS ACTIVITY
invoke
mechanism
trigger
references
referenced by
BUSINESS EVENT
PEOPLE
provides data to
trigger
BUSINESS PROCESS
invoke
provides data to
invokes
DATA ACCESS PROCESS
DATA MODEL COMPONENTS
manipulate
invoke
90
Levels of Process Modeling
Logical Data Process
start
Activity Decomposition
Phase 2 Data Model
Position Management
Accept PRODUCT
no more
continue
found
not found
Read SUPPLIER
Find OTHER
Develop job description
Compare applicant skills to job
Increase job skill levels
none found
found
none chosen
Choose SUPPLIER
chosen
Workflow
Present SUPPLIER
Receive Material Request
Check Warehouse Stock Level
ok
Start
Select Supplier
Evaluate Parts Use
stop
91
Activity Based CostingIntegrated With Activity
Modeling
Business Rules
To-Be IDEF0
As-Is IDEF0
Financial Analysis
ABC
Improvement
Diagnosis
92
Activity Modeling
Activities are viewed in the detailed diagram or
the hierarchy (equivalent to node
tree). Modifications to either view should
automatically update the other.
93
Business Rule Life Cycle
PLANNING REQUIREMENTS
StrategicDirection
InformationRequirement
Technology Independent
InformationUsage
Technology/SystemRequirements
Technology Dependent
PerformanceRequirement
SYSTEM IMPLENTATION
94
Database Design
95
Data Warehouse Design
96
Logical Model to Physical Design Matrix(one to
many)
This Logical to Physical View matrix graphically
shows the relationship between logical model
attributes and physical design data items.
97
Data Model to Design Cross Reference
The Design Object Person implements attributes
from the Entities Person and Employee. You must
be able to optimize your physical model to your
particular hardware and software constraints,
without changing or compromising the rules
expressed in the logical model.
98
Design to Data Model Cross Reference
The Entity Person is implemented in the Pay
System, the HR system and the Pers Data
system. Duplication is identified for elimination
or synchronization.
99
Database and System Design Technology-
Dependence User-Defined Objects
Tables, fields, relationships, and referential
integrity (PKey/FKey) characteristics can be
normalized differently than the data model for
performance needs. Design should be available in
text or diagram form. Design may be extended to
include additional objects such as menus,
screens, dialogs, and more.
100
Business Rule Life Cycle
PLANNING REQUIREMENTS
StrategicDirection
InformationRequirement
Technology Independent
InformationUsage
Technology/SystemRequirements
Technology Dependent
PerformanceRequirement
SYSTEM IMPLENTATION
SYSTEM IMPLENTATION
101
Generate DDL
102
SQL (DDL) Schema Report
103
Generate XML
104
XML Schema Report
105
Business Rules Meta-Data

• Allows Maximum Flexibility
• Reduces System Maintenance
• Simplifies Design and Implementation
• Ensures that systems fully support business
  needs
• MIS personnel don't need to be business experts
• Rules can change without affecting
  implementation effort

• Benefits of Managing Business Rules as Meta-Data
• Techniques for Defining, Documenting, Modeling,
  and Implementing Business Rules
• The Business Rule Life Cycle

• Business Statements
• Data Entities and Attributes
• Entity Associations
• Domain Values
• Derivations
• Rule Expressions

• Strategically Driven
• Information Centric
• Customer Focused
• Model Based
• Disciplined
• Technology Independent

106
Who are you going to call?
Alan PerkinsVice President, Client
ServicesVisible Systems Corporation 901 North
Washington StreetAlexandria, Virginia
22314 800-723-8924 ext. 225 fax
703-739-0074 alan.perkins_at_visible.com http//www.v
isible.com