Introduction to Decision Model and Notation (DMN™)

What is DMN?

Decision Model and Notation (DMN) is an industry standard created by the OMG® (OMG) that provides a common notation for decision modeling. It bridges the gap between business decision design and decision implementation by providing a notation that's readable by both business and technical users.

DMN is designed to:

• Document decision-making processes clearly and precisely
• Enable collaboration between business and technical teams
• Create executable decision models that can be automated
• Ensure decision logic is testable, auditable, and traceable

Core Components of DMN

DMN models consist of two key elements:

1. Decision Requirements Diagram (DRD) - The visual representation of the decision model
2. Decision Logic - The detailed specifications of how decisions are made

Decision Requirements Diagram (DRD)

A DRD shows the structure of a decision model using these elements:

• Decision Nodes (rectangles): Represent the decisions to be made
• Input Data (ovals): Represent information used to make decisions
• Knowledge Sources (documents): Represent sources of knowledge (policies, regulations, etc.)
• Business Knowledge Models (rounded rectangles): Represent reusable decision logic
• Information Requirements (solid arrows): Show dependencies between decisions
• Knowledge Requirements (dashed arrows): Show dependencies on business knowledge models
• Authority Requirements (dotted arrows): Show dependencies on knowledge sources

Decision Logic

Decision logic in DMN can be expressed in several ways:

1. Decision Tables: Structured way to represent complex decision logic
2. Literal Expressions: Simple expressions or calculations
3. Function Definitions: Reusable logic functions
4. Invocations: Calls to Business Knowledge Models
5. Contexts: Collections of local variables and results
6. Lists: Collections of items

The most common format is the decision table, which represents conditions and actions in a tabular format.

Decision Tables

Decision tables are a powerful way to express complex business rules. They consist of:

• Hit Policy: Determines how rules are evaluated (unique, first, collect, etc.)
• Input Columns: Conditions to be checked
• Output Columns: Actions or results
• Rules: Rows that connect inputs to outputs

For example, a basic decision table might look like this:

U	Age	Years of Service	Vacation Days	Annotation
	<18	-	5	Minor employees
	>=18	<5	10	Standard entry level
	>=18	[5..10]	15	Experienced employees
	>=18	>10	20	Senior employees

This table uses a Unique (U) hit policy, meaning only one rule will match for any given input.

FEEL - Friendly Enough Expression Language

DMN uses a language called FEEL (Friendly Enough Expression Language) for expressing decision logic. FEEL is designed to be:

• Readable by business users
• Precise enough for execution
• Capable of expressing complex logic

Examples of FEEL expressions:

• Simple calculations: Base Premium * Risk Factor + 10
• Conditional logic: if Age < 25 then "Young Driver" else "Regular Driver"
• List operations: some item in [1, 2, 3] satisfies item > 2
• Date handling: date("2023-04-15") + duration("P3M") (adds 3 months to the date)

DMN Execution

DMN models are not just documentation—they're executable specifications. When a DMN model is executed:

1. Input data is provided to the model
2. The model evaluates decisions in dependency order
3. Decision logic is applied according to the specified rules
4. Results are returned for all requested decisions

This execution can happen within applications, as standalone services, or as part of larger process workflows.

Benefits of Using DMN

DMN offers numerous advantages over traditional approaches to defining business rules:

1. Business-IT Alignment: Common language for business and technical stakeholders
2. Separation of Concerns: Decision logic separated from process flow
3. Transparency: Clear documentation of decision criteria and logic
4. Consistency: Standard approach to modeling decisions across the organization
5. Modularity: Reusable decision components
6. Testability: Easy to validate with test scenarios
7. Adaptability: Easier to change decision logic without affecting applications
8. Interoperability: Standard format works across compliant tools and engines

DMN in the Apache KIE™ Ecosystem

Within the Apache KIE™ ecosystem, DMN is fully supported with these features:

• Visual modeling through Aletyx Playground
• Full compliance with the DMN specification
• Integration with other standards like BPMN
• Cloud-native deployment options
• Testing and validation tools
• Versioning and governance capabilities

Getting Started with DMN

The best way to learn DMN is by creating your first model. In the next guide, Your First DMN Model, we'll walk through creating a simple vacation days calculation model that demonstrates the key concepts of DMN.

Ready to create your first DMN model? Let's get started!

PREVIOUS Decision Model and Notation Overview

NEXT Basic DMN Examples

Was this page helpful?