Decision Model and Notation

Decision Model and Notation
Name	Decision Model and Notation
Othernames	DMN
Developer	Object Management Group
FirstReleased	2015
LatestRelease	1.3
Status	Standard
Website	Object Management Group

Decision Model and Notation is a standardized notation and metamodel for modeling and executing business decisions, developed to interoperate with process and rules standards. It provides a shared language for analysts, architects, developers, and auditors to represent decision logic in tabular and graphical forms that can be executed by compliant engines. The notation integrates with other standards and technologies to support traceability, governance, and automation across enterprise platforms.

Overview

Decision Model and Notation defines a visual and XML-based representation for decision models that separates decision logic from process orchestration. It complements standards such as Business Process Model and Notation, Unified Modeling Language, XML Schema, SOSA, and aligns with governance frameworks used by European Commission, US General Services Administration, World Bank, International Organization for Standardization, and Institute of Electrical and Electronics Engineers. The notation emphasizes clarity and testability to enable collaboration among stakeholders including teams from Accenture, IBM, Red Hat, SAP SE, and Oracle Corporation.

History and Development

DMN originated from industry demand for a vendor-neutral decision standard and was advanced through technical committees led by the Object Management Group, with contributions from representatives of Lloyd's of London, Cambridge University, University of Oxford, MIT, Gartner, Forrester Research, and vendor consortia. The initial release followed iterative public reviews involving standards bodies such as World Wide Web Consortium, OASIS, European Telecommunications Standards Institute, and government pilot projects like those run by the UK Cabinet Office and US Department of Defense. Subsequent revisions incorporated feedback from practitioners at PwC, Deloitte, KPMG, McKinsey & Company, and open-source communities including Apache Software Foundation and Eclipse Foundation.

Specification and Components

The specification defines core constructs: the Decision Requirements Diagram, decision tables, boxed expressions, and an interchange format in XML and JSON aligned with XML Schema and JSON Schema. It also specifies a standardized expression language and type system that interoperates with FEEL (Friendly Enough Expression Language), Java, JavaScript, Python, and rule engines like those from Drools, IBM Operational Decision Manager, and Red Hat Decision Manager. Governance features map to artifacts used by International Electrotechnical Commission standards and audit logs familiar to practitioners at Goldman Sachs, JP Morgan Chase, European Central Bank, and Federal Reserve Board.

Modeling Concepts and Notation

Core modeling artifacts include Decision Requirements Diagrams linking decisions, input data, knowledge sources, and business knowledge models; decision tables that enumerate rules and hit policies; boxed expressions for literal logic; and annotations for business knowledge models tied to policies or regulations. The notation supports expression languages such as FEEL, as well as embedding expressions in XPath, SQL, JavaScript, and C# to interoperate with enterprise stacks like those at Amazon Web Services, Microsoft Azure, Google Cloud Platform, Salesforce, and ServiceNow. The Decision Requirements Diagram is often used in tandem with process models created by teams at Siemens, General Electric, Boeing, and Airbus for regulated workflows.

Execution and Implementation

Execution semantics specify how decision graphs evaluate, including dependency resolution, data coercion, and hit policies that determine conflict resolution. Implementations exist as engines and runtime components that integrate with orchestration platforms such as Camunda, Kubernetes, Apache Kafka, and Spring Framework. Commercial implementations are provided by IBM, Red Hat, TIBCO Software, FICO, and SAP, while open-source projects at GitHub and contributions from Eclipse Foundation provide SDKs and reference runtimes. Integration patterns include model-driven deployment pipelines used by Atlassian, GitLab, Jenkins, and CircleCI.

Use Cases and Applications

DMN is applied in domains requiring transparent, auditable decision logic: credit adjudication at Wells Fargo and Bank of America, insurance underwriting at AXA and Allianz, compliance and tax determination in projects at HM Revenue and Customs, Internal Revenue Service, and European Banking Authority, and clinical decision support in initiatives associated with World Health Organization, National Health Service (England), and academic hospitals such as Mayo Clinic and Johns Hopkins Hospital. Other applications include supply chain optimization at Procter & Gamble, Unilever, automated pricing at eBay, Amazon (company), and policy enforcement in telecommunications by Verizon Communications and AT&T.

Tools and Industry Adoption

A broad ecosystem supports authoring, testing, and governance: commercial editors from Trisotech, Signavio, Sapiens, and Decisions; execution engines from Drools, IBM ODM, Red Hat Decision Manager; and collaborative platforms like Confluence, JIRA (software), Microsoft Teams, and Slack (software). Certification programs and conformance test suites coordinated by the Object Management Group facilitate adoption among enterprises such as Facebook, Twitter, LinkedIn, and TikTok (service). Academic courses and professional training are provided by Harvard University, Stanford University, Carnegie Mellon University, and Columbia University to support practitioners entering sectors led by Siemens Healthineers, Philips, and Roche.

Category:Standards