the decision model - dama ny€¦ · method decision model notation conditions conclusion pattern...
TRANSCRIPT
The Decision Model: Its Debut at MIT’s IQ Symposium as the Basis
for an IQ Framework
Barbara von HalleManaging Partner
Knowledge Partners International, LLC2010
About KPI
Copyright © KPI 2010 2
Publications
Thought Leader
The Decision ModelBusiness Logic Framework linking
Business with Technology
Business Process ManagementBusiness Decision Management
Business Rule ManagementEnterprise Architecture
Business AnalysisRequirements
Testing
Services
FirstSTEPService to create
unambiguous, and complete Requirements
KPISTEPService to perceive, organize
and manage Business Processes and Rules with
Decision Models
STEPmentMentoring of clients to
achieve self-reliance with Center of Excellence
Training & Certification
Experience
Financial Services
Insurance
Healthcare
Government
Utilities
Transportation
Telecommunication
Energy
Copyright © KPI 2010 3
“..one of the classic books
of a new era in computing
that will have much traction
in the next few years”Dr. Opher Etzion,
Master Inventor, IBM
Agenda
• Motivation for The Decision Model
• Introduction To The Decision Model
• Top Down, Iterative, Agile
• Impact on Business Process
• Recent Advances and Real World Experiences
• The Decision Model and Information Quality
• Summary
Copyright © KPI 2010 4
Separation of Concerns:The One Dimension Left Behind
5
Security Component
Workflow Component
Transaction Component
Presentation Component
Base Application
Business Logic
ComponentReporting/BI Component
Database Component
A “Big Ball of Mud”(Foote & Yoder)
Component-Based Architecture (Orr)
Copyright © KPI 2010
How Business Rules (Logic) are Handled Today
66Copyright © KPI 2010
Copyright © KPI 2010 7
How The Decision Model Changes Everything
Process Model Decision
Model
Rule Families
Agenda
• Motivation for The Decision Model
• Introduction To The Decision Model
• Top Down, Iterative, Agile
• Impact on Business Process
• Recent Advances and Real World Experiences
• The Decision Model and Information Quality
• Summary
Copyright © KPI 2010 8
Definition of Business Logic
9
Business Logic is the means by which the business
derives conclusions from facts.
Copyright © KPI 2010
The simplest case is the evaluation of single fact, leading to a single conclusion:
One example of such a statement:
A person is highly likely to default on a loan
A person has a poor employment history
Let us examine “Fact”
• Fact Type is the same as an attribute type
• Fact Value is the same as an attribute value
– Person First Name
• Fact Type
– John Doe
• Fact Value
10Copyright © 2010 KPI
Domains
• Fact Type Domain is the same as Attribute Type Domain.
• Fact Type Domains are important to the integrity of the business logic
• How so?
11
Fact Type Fact Type Domain
Person Employment History Good, Average, Poor
Person Likelihood of Default on Loan High, Medium, Low
Claim Payment Amount Min Claim Amount to Max Claim Amount
Policy Renewal Method Manual, Automatic
Copyright © 2010 KPI
What is an Atomic Piece of Business Logic?
• An atomic piece of business logic
– Consists of zero to many conditions
– Leading to a conclusion about one fact type
– Each condition is an atomic logical expression
– About an atomic fact type
– Conditions are ANDed together, never ORed
12Copyright © KPI 2010
The Rule Family –A Way to Represent Multiple Logic Statements
13
Conditions Conclusion
Person Employment
History
Person Mortgage
Situation
Person Miscellaneous
Loans Assessment
Person Likelihood of
Defaulting on a Loan
Is Poor Is Poor Is High Is High
Is Good Is Low
Is Poor Is Poor Is Low Is Medium
PersonEmployment History
Is Poor PersonMortgageSituation
Is Poor PersonMisc Loan Assessment
Is HighAND
AND
PersonLikelihood of Defaulting on a Loan
Is High
Instead of Multiple Logic Statements that Look Like This:
They May be Represented in Two Dimensional Tables called Rule Families:
Rule Families are Tables that Conform to Rigorous Principles
Copyright © KPI 2010
Building Further: Where Do We Get Our Input?
• Starting with the first condition, we ask where this Fact value comes from. Input from a web page or a file? Is it Persistent data? Is it the result of execution logic?
• In this case we discover that it comes from executing logic that evaluates other business criteria: the business experts want to judge a Person’s Employment History based on criteria such as Person’s Years at Current Employer and Person’s Number of Jobs in the Past Five Years.
• We have to build an additional Rule Family where the conclusion will be “Person Employment History”, a different conclusion to that of our current Rule Family (Rule Family: Business logic grouped by Conclusion.)
14
Conditions Conclusion
Rule
Pattern
Person
Employment
History
Person
Mortgage
Situation
Person
Miscellaneous
Loans
Assessment
Person
Outside
Credit
Rating
Person Likelihood
of Defaulting on a
Loan
1 is Poor Is Poor Is High is High
Copyright © KPI 2010
Building Up to Two Rule Families
• Note the Interim Conclusion “Person Employment History”• We discover the need for yet another Rule Family. This one comes
to a conclusion about a Person’s Employment History which is based on two conditions: Person Years at Current Employer and Person Number of Jobs in Past Five Years.
15
Conditions Conclusion
Rule
Pattern
Person
Employment
History
Person
Mortgage
Situation
Person
Miscellaneous
Loans Assessment
Person
Outside
Credit Rating
Person Likelihood of
Defaulting on a Loan
1 is Poor Is Poor is High ? ? is High
Conditions Conclusion
Rule
Pattern
Person Years at Current
Employer
Person Number of Jobs in
Past Five Years
Person Employment History
Copyright © KPI 2010
Dynamic Data
What is it?
• Created on the fly
• Whenever the business comes to a conclusion even if the conclusion is not stored
• Conclusions guide transactions and business processes
• Conclusions determine outcome, success or failure of a transaction, process, or decision
Why is it important?
• It runs the business!
• Discovering it and managing it instigates business innovation.
• This can make a huge difference in organizational maturity, performance, perhaps survival.
• The data professional brings good data practices to the management of dynamic data.
16Copyright © 2010 KPI
Decision Model Principles
• Structural Principles – Structural simplicity
• Declarative Principles – Declarative structure
• Integrity Principles – Optimal logical integrity
Copyright © KPI 2010 17
These Principles ensure that each row, each pattern and each family has business and logical integrity: this
means that the business purpose has been understood and aligned, and that there is no logical error in the
logic, and that there is no conflict or duplication in the logic. The Principles introduce Normalization.
Decision Model Normalization
• Normalization: the concept of decomposing structures into more desirable structures that have greater levels of integrity– First normal form – ensures business logic represented and
interpreted in only one way as a Rule Family with one conclusion fact type and conditions ANDed
– Second normal form – eliminates irrelevant conditions
– Third normal form – eliminates hidden Rule Families in conditions
Copyright © KPI 2010 18
How Should We Separate the Decision Dimension?
• Universal Model for Decision Structures (Decision Model)
– Normalized Logic Entities
– Inferential relationships
– Logic statements (content)
• Decomposition into Atomicity of Logic
• Normal Forms
• Best Practices
– Naming Conventions, definitions
– Domains
– Not overloading
– Role of nulls
– Integrity Validation
• Modeling tools and generation of code
Copyright © 2010 KPI 19
Decision
Business
Motivation
Model
Decision Model:business rules and
business logic
Process Model
Use Cases
SOA Components
Business
Requirements &
Test Cases
Vocabulary Models:
Glossary/Semantic Model
Logical Data Model
Object Model
Enterprise
Business Unit
Function Function
Business Unit
Function
Organization
Model
Agenda
• Motivation for The Decision Model
• Introduction To The Decision Model
• Top Down, Iterative, Agile
• Impact on Business Process
• Recent Advances and Real World Experiences
• The Decision Model and Information Quality
• Summary
Copyright © KPI 2010 21
Every Decision Model Starts with a Business Decision
22
“Business decision: a conclusion that a business arrives at through business logic and which the business is interested in managing.”
22
The underlined words (Calculate, Estimate, Determine, Assess, Validate) are “Decision Words”
Fact Type Business Decision
Claim Payment Amount Estimate the claim payment amount
Claim Payment Eligibility Determine Claim Payment Eligibility
Customer Likelihood of Loan Default Determine Customer Likelihood of Loan Default
Insurance Policy Renewal Method Determine insurance policy renewal method
Inventory Item Minimum Stock Level Assess the Inventory Item minimum stock level
Loan Prequalification Determine loan prequalification requirements for a customer
Person BMI (Body Mass Index) Calculate Person BMI
Vendor Performance Index Calculate the Vendor Performance Index
Copyright © KPI 2010
Determine Policy
Renewal Method Decision Model
Notation
23Copyright © KPI 2010
Determine Policy
Renewal Method Decision Model
Notation
24Copyright © KPI 2010
Policy Renewal MethodPolicy Pricing Within Bounds
Policy Underwriting RiskManual Underwriting Indicator
Policy Renewal MethodPolicy Pricing Within Bounds
Policy Underwriting RiskManual Underwriting Indicator
Policy Pricing Within BoundsPolicy Discount
Policy Tier
Determine Policy
Renewal Method Decision Model
Notation
25Copyright © KPI 2010
Policy Renewal MethodPolicy Pricing Within Bounds
Policy Underwriting RiskManual Underwriting Indicator
Policy Pricing Within BoundsPolicy Discount
Policy Tier
Determine Policy
Renewal Method Decision Model
Notation
Conditions Conclusion
Pattern Policy Tier Policy Discount
Policy Pricing Within Bounds
1 ≤ 1 Is No
2 ≤ 1.5 > 10% Is No2 ≤ 2 > 20% Is No2 ≤ 2.6 > 22% Is No2 > 1 ≤ 0% Is Yes
2 > 1.5 ≤ 20% Is Yes2 > 2 ≤ 22 Is Yes
1 > 2.6 Is Yes
Conditions Conclusion
PatternPolicy Pricing Within Bounds Policy Underwriting Risk
Manual Underwriting Indicator Policy Renewal Method
1 Is Unacceptable Is Manual Renewal Process
2 Is No Is Manual Renewal Process
3 Is On Is Manual Renewal Process
4 Is Yes Is Acceptable is Off Is Automatic Renewal Process
26Copyright © KPI 2010
Policy Underwriting RiskInsured Major Ownership Change
Insured Major Location Change Policy Annual Premium
Policy Discontinued Agent
Insured Major Ownership Change
Insured Minority StockholderInsured Majority Stockholder
Insured Board ChangeInsured CEO Change
Insured Major Location Change
Insured Location Zip-5 Insured Location Occupied Square Footage
Insured Location Construction
Policy Discount
Policy Grade Package Grade
Package Discount Location State Category
Decision Model Notation
27Copyright © 2010 KPI
Policy Renewal MethodPolicy Pricing Within Bounds
Policy Underwriting RiskManual Underwriting Indicator
Policy Pricing Within BoundsPolicy Discount
Policy Tier
Determine Policy
Renewal Method
Agenda
• Motivation for The Decision Model
• Introduction To The Decision Model
• Top Down, Iterative, Agile
• Impact on Business Process
• Recent Advances and Real World Experiences
• The Decision Model and Information Quality
• Summary
Copyright © KPI 2010 28
Option 1
Option 2
Option 3
Rule
Pattern
1 is Low is Good = "A"
1 is Low is Bad = ?
1 is High is Good = ?
1 is High is Bad = ?
Conditions Conclusion
Person's Debt
Person's Employment
History
Person's Credit
Rating
Process Model Rule Family Table Decision Model Diagram
Distinguishing Decisions from Process
29Copyright © KPI 2010
Simplify the Models, Improve the Solution
Before After
3030Copyright © KPI 2010
Agenda
• Motivation for The Decision Model
• Introduction To The Decision Model
• Top Down, Iterative, Agile
• Impact on Business Process
• Recent Advances and Real World Experiences
• The Decision Model and Information Quality
• Summary
Copyright © KPI 2010 31
Recent Advances in the Model and Practice
• Decision Model Views
– Mass Customization: Geographical, political, glossary, customer-based
• Codeless implementation into BRMS
• The Decision Model Data Quality Framework
• The Decision Model for Data Translation Services
• Testing and Generation of Test Cases
Copyright © KPI 2010 32
Real World
• “The Decision Model’s principles and normalization rules give us confidence we can get repeatability and consistency amongst business analysts when performing rules analysis.
• In addition, the structural integrity of the Decision Model makes the technology implementation straightforward
• IT and Operations have agreed to use our decision model as business requirements for business logic changes – this will greatly speed up the change process
• In addition, the use of a COTS BRMS solution will allow us to take advantage of additional capabilities over time, such as enhanced testing and decision-warehousing capabilities.“
• From policy to automation reduced by 30% in time, while delivering 66% more changes
Mark Pettit, Freddie Mac, Operations Management Group, MIT IQIS, July 15, 2010
Copyright © KPI 2010 33
Agenda
• Motivation for The Decision Model
• Introduction To The Decision Model
• Top Down, Iterative, Agile
• Impact on Business Process
• Recent Advances and Real World Experiences
• The Decision Model and Information Quality
• Summary
Copyright © KPI 2010 34
The Decision Model IQ Framework
35
A Decision Model Diagram for Completeness
Copyright © 2010 KPI 36
Rule Family for Completeness
Rule Pattern Conditions Conclusion
Applicant Annual Salary
Documentation
Applicant Supplemental
Income Amount
Applicant Supplemental
Income Amount Completeness
Is Acceptable Is Complete
Is Not
acceptable
Is Populated Is Complete
Is Not
acceptable
Is Not
populated
Is Not complete
Copyright © 2010 KPI 37
Other IQ Dimensions
• Consistency, reasonableness, and accuracy logic originate from– Analysis of internal and external data– Insights from vendors– Fraud detection programs– Analytics– Organizational experience over time
• Because these three dimensions involve conditions leading to a conclusion, the Decision Model View is an appropriate vehicle for managing these specifications.
Copyright © 2010 KPI 38
Agenda
• Motivation for The Decision Model
• Introduction To The Decision Model
• Top Down, Iterative, Agile
• Impact on Business Process
• Recent Advances and Real World Experiences
• The Decision Model and Information Quality
• Summary
Copyright © KPI 2010 39
Five Reasons to Include the Decision Model
1. The business rules of business decisions are the forgotten dimension.
2. The Decision Model improves business decision-making.
3. The Decision Model transcends and leverages available data.
4. The Decision Model is iterative, agile in nature.5. The Decision Model always reveals important
pieces of information we have ignored (dynamic data).
40Copyright © 2010 KPI
Five Reasons to Manage Dynamic Data
1. Dynamic data is the glue that ties together the way the business thinks and determines the quality of its decisions.
2. Dynamic data emerges as an asset as important as persistent data.3. If dynamic data is not managed property, Decision Models fall apart.4. If Decision Models fall apart, business conclusions may be risky.
Risky conclusions result from:• Incomplete domain coverage• Overloaded conclusion values• Conclusion values that are null
5. If business conclusions are risky, the business may suffer.
The business needs data professionals to help!
41Copyright © 2010 KPI
Next Steps
Copyright © KPI 2010 42
Visit our web siteswww.thedecisionmodel.comwww.kpiusa.com
• Updated Events• Download White Papers• News
Become a member at Linkedin The Decision Model Group
• Network with practitioners• Exchange ideas• Learn new tricks
Read our articles and bookswww.modernanalyst.comwww.tdan.comwww.requirementsnetwork.com
Join our presentationsWe conduct frequently presentations for following interest groups:
www.theiiba.orgwww.bpminstitute.comwww.dama.org
Contact uswww.Twitter.com/[email protected] +1 973 543-1339
Start a projectDiscuss with us how to apply The Decision Model for your Requirement or Business Rules project.