Knowledge Modeling Analysis method

Shape-First Modeling

Designs RDF data requirements first as SHACL shapes and validates sample data early.

SHACL separates the shapes graph from the data graph. NodeShapes select targets; PropertyShapes define path, cardinality, datatype, class, and other constraints. Validation produces a results graph.

Core question

Which RDF structure must hold for the use case, and how is each violation reported?

Plan Shape-first Modeling Method guide

Start comparison

Give feedbackIs this method useful to you?

Participants: 1-4
Time: Halber Tag pro Domain Slice

Shape-first Modeling uses targets, NodeShapes, PropertyShapes, and constraint components before broad ontology modelling. Assessments remain justified and context-bound.

Paper illustration of Shape-first Modeling with its method-specific working model.

Define contract
Define contract
Define use case, target, must/should rules, and severity.
Author shapes
Author shapes
Write NodeShapes and PropertyShapes with Core components.
Build fixtures
Build fixtures
Create valid, invalid, and boundary data.
Validate
Validate
Run a SHACL processor and inspect the results graph.

Update after new evidence or a material context change and retain the previous version.

A good fit when

For RDF interfaces and knowledge graphs with concrete data quality contracts.

Ideal for

Data quality in knowledge graphs
Linked data validation
Contract-based interfaces between producer and consumer
GraphRAG quality gates

Not good for

Very exploratory sketching phases
Purely schema-on-read applications without a quality requirement

Outputs

SHACL shape suite
SHACL shape suite with traceable sources, assumptions, and domain decisions.
Decision and review trail
Owners, dates, and next review for SHACL shape suite.

SHACL shape suite

Fictional worked example

targets, NodeShapes, PropertyShapes, and constraint components before broad ontology modelling as a precise working structure with a fictional example.

Target

NodeShape · targetClass/node

ActionShape targetClass Action.

Property shape

Path · min/maxCount · datatype/class

owner min=1 max=1; dueDate xsd:date.

Fixture

Positive · negative · boundary

Action without owner as negative case.

Result

focusNode · resultPath · severity · message

A-17, owner, Violation, owner missing.

The example demonstrates the domain logic of Shape-first Modeling; worksheets remain blank.

W3C Shapes Constraint Language (SHACL)

First step

Start with the core question: Which RDF structure must hold for the use case, and how is each violation reported?

Open the method guide

Good to do first

Knowledge Modeling

Competency Questions

Supplies RDF use cases from which testable shapes are derived.

Knowledge Modeling

Property Graph Schema Design

Supplies constraints that are formalized here as SHACL or ShEx rules.

Where to go next?

Knowledge Modeling

Ontology Design Patterns

Checks whether validated data structures satisfy existing design patterns.

Knowledge Modeling

Linked Open Terms

Publishes the validated structures as part of an open vocabulary.

Deep dive

Purpose

Shape-First Modeling defines data quality through shapes before implementation or integration frays at the edges. The approach fits when validation and data contracts should be part of the design from the start.

How it works

First, the relevant node classes and mandatory properties are described. The team then translates these requirements into SHACL or ShEx, validates sample data, and refines the rules afterward. In the end, checkable shapes remain that can be transferred directly into pipelines and contracts.

In detail

Shape-First Modeling works because data quality is fixed already in the model, not only after a flawed integration. The approach gains sharpness when shapes are derived from real data contracts, mandatory statements, and validation rules. Especially important are understandable error messages, realistic sample data, and clean versioning, so that validation does not slow things down but enables reliable collaboration between sources.

Facilitation

A small, business-critical data exchange, where the benefit and limits of the rules become visible quickly, is suitable for introduction. The tipping point usually lies with shapes that are too strict and blank out real data situations, or with rules too soft to offer protection. Close with a working validation path, a few prioritized shapes, and an agreement on how future changes will be introduced.

The method guide walks execution through 4 phases, timing, 4 pitfalls, and clear stop criteria.

Open method guide

Your perspective

Does this method help with your work?

Your feedback helps shape the next improvements to methodatlas.

Sources and topics

Topics

Knowledge graph Validation Semantic

Sources

W3C Shapes Constraint Language (SHACL)w3.orgW3C recommendation for SHACL; defines node and property shapes for validating RDF graphs.

In playbooks

Prepare Knowledge ModelingMove from a knowledge domain to questions, concept model, schema, and validation rules.

Artifact templates

Shape-First Modeling: WorksheetPlan Shape-first Modeling around its domain steps and the concrete decision question.

markdown

shape-first-modeling-working-template.md

Plan Shape-first Modeling around its domain steps and the concrete decision question.

Shape-First Modeling: Worksheet

Question

To be confirmed

Desired outcome

To be confirmed

Scope

Full workshop Time: 60–120 minutes per shape module

Complete this worksheet on paper or in your own document during the work.

Preparation for this scope

Method setup

Materials: RDF use cases, valid and invalid sample data, namespace and severity policy, shared workspace, source links, and decision log.
Roles: Domain expert · RDF engineer · data producer · consumer
Advance information: Collect RDF use cases, valid and invalid sample data, namespace and severity policy in advance and mark open assumptions.
Overall time needed: 60–120 minutes per shape module
Setup: Prepare targets, NodeShapes, PropertyShapes, and constraint components before broad ontology modelling visibly and calibrate assessment terms before starting.

Prepare for the work steps

Define contract: Define contract

RDF use cases
valid and invalid sample data
namespace and severity policy

Author shapes: Author shapes

constraint backlog
valid and invalid sample data
namespace and severity policy

Build fixtures: Build fixtures

shapes graph
valid and invalid sample data
namespace and severity policy

Validate: Validate

RDF fixtures
valid and invalid sample data
namespace and severity policy

SHACL shape suite

Target

NodeShape · targetClass/node

...

Property shape

Path · min/maxCount · datatype/class

...

Fixture

Positive · negative · boundary

...

Result

focusNode · resultPath · severity · message

...

Enter only real input, sources, and decisions.

W3C Shapes Constraint Language (SHACL)

Define contract: Define contract

Expected artifact: constraint backlog

constraint backlog · source · uncertainty · next decision

Entry:

...

Is every rule justified by the domain?

Author shapes: Author shapes

Expected artifact: shapes graph

shapes graph · source · uncertainty · next decision

Entry:

...

Are paths, counts, datatypes, and classes correct?

Build fixtures: Build fixtures

Expected artifact: RDF fixtures

RDF fixtures · source · uncertainty · next decision

Entry:

...

Does each negative case violate the intended rule?

Validate: Validate

Expected artifact: validation suite

validation suite · source · uncertainty · next decision

Entry:

...

Does every result include focus node, path, and constraint?

Open questions and next steps

...

Method guide: https://methodatlas.meierhoff-systems.de/en/methods/shape-first-modeling/run-sheet

Alternatives and neighbors

Fitting alternatives with decision guidance, followed by other related methods.

All methods

Knowledge Modeling

Property Graph Schema Design

Instead of Shape-First Modeling, choose Property Graph Schema Design when you want to align a graph schema directly with queries, traversals, and edge logic.