Skip to content

data-modelling.md

Latest commit

 

History

History
195 lines (140 loc) · 9.72 KB

data-modelling.md

File metadata and controls

195 lines (140 loc) · 9.72 KB
Raw

Solid Data Modelling Skill

You are an expert on data modelling for the Solid ecosystem. This includes FAIR data principles, RDF vocabularies and ontologies and SHACL shapes.

FAIR Data Principles in Solid

FAIR Data Principles (Findable, Accessible, Interoperable, Reusable) describe guidelines for improving data usability in open and distributed systems. In the context of Solid data modelling, implementing FAIR principles ensures data can be easily discovered, accessed, combined, understood, and reused across applications and domains.

Principle Solid Implementation
Findable Resources have dereferenceable HTTP URIs; The Solid Type Index provides a standard machine-readable discovery mechanism to enable data discovery in a Solid pod
Accessible Resources are served over standard HTTP with authentication (Solid-OIDC) and authorisation (ACP/WAC)
Interoperable Data uses RDF with standard vocabularies (schema.org, FOAF, LDP, etc.)
Reusable Shapes (SHACL) define reusable data schemas; linked data enables cross-app data reuse

Applying FAIR in Practice

  1. Mint stable URIs — use consistent URI patterns for resources
  2. Use standard vocabularies — reuse standard vocabularies e.g. schema.org, FOAF, Dublin Core, SKOS where possible
  3. Describe your data in machine-readable format — add RDF type statements and metadata to every resource
  4. Publish shapes — provide SHACL shapes so consuming apps can validate the data
  5. Link datasets — use rdfs:seeAlso, and skos:exactMatch to connect related data

RDF Vocabulary Selection

Core Vocabularies for Solid

Prefix Vocabulary Description Namespace URL
ACP acp: Access Control Policy Language http://www.w3.org/ns/solid/acp#
AS (ActivityStreams) as: Social activities, notifications https://www.w3.org/ns/activitystreams#
DC dc: / dcterms: Metadata (title, creator, date, description) http://purl.org/dc/terms/
FOAF foaf: Person, identity, social graph http://xmlns.com/foaf/0.1/
ICAL ical: Calendar data http://www.w3.org/2002/12/cal/ical#
LDP ldp: Containers, membership http://www.w3.org/ns/ldp#
ORG org: Organization ontology http://www.w3.org/ns/org#
OWL owl: Ontology relations, sameAs http://www.w3.org/2002/07/owl#
PIM pim: Storage, preferences, configuration http://www.w3.org/ns/pim/space#
POSIX posix: POSIX file system permissions http://www.w3.org/ns/posix/stat#
RDF rdf: RDF syntax concepts http://www.w3.org/1999/02/22-rdf-syntax-ns#
RDFS rdfs: Labels, comments, subclass relationships http://www.w3.org/2000/01/rdf-schema#
Schema.org schema: Structured data vocabulary https://schema.org/
Solid solid: OIDC, notifications, identity https://www.w3.org/ns/solid/
vCard vcard: Contact information http://www.w3.org/2006/vcard/ns#

Vocabulary Lookup

Before creating custom RDF classes and properties, search existing ontologies/vocabularies:

Shapes: SHACL

Shapes define the expected structure of RDF data. They serve as schemas for validation and code generation.

SHACL (Shapes Constraint Language)

SHACL is the W3C Recommendation for RDF validation, more expressive for complex constraints.

@prefix sh: <http://www.w3.org/ns/shacl#>.
@prefix foaf: <http://xmlns.com/foaf/0.1/>.

<PersonShape>
    a sh:NodeShape ;
    sh:targetClass foaf:Person ;
    sh:property [
        sh:path foaf:name ;
        sh:datatype xsd:string ;
        sh:minCount 1 ;
    ] ;
    sh:property [
        sh:path foaf:mbox ;
        sh:nodeKind sh:IRI ;
        sh:maxCount 1 ;
    ] .

SHACL Shape Discovery

A number of public SHACL shapes collections and published shapes are available that you can reuse or study:

MCP Adapters for Ontology Discovery

Model Context Protocol (MCP) adapters allow AI tools and development environments to discover and work with ontologies and shapes programmatically.

Discovering Ontologies via MCP

When building an MCP adapter for ontology discovery, key capabilities to expose:

  1. Search by term — find classes and properties matching a keyword
  2. Shape retrieval — fetch SHACL shapes for a known type
  3. Type index traversal — discover where a user stores instances of a given type on their pod

Recommended Approach for Ontology Discovery in Apps

1. Check LOV (lov.linkeddata.es) for existing vocabulary terms
2. Use SHACL shapes (from Solid shapes repo) to validate data
3. For pod-specific discovery, traverse the pod's Type Index

Structure

# Public type index at /settings/publicTypeIndex.ttl

@prefix solid: <http://www.w3.org/ns/solid/terms#>.
@prefix schema: <https://schema.org/>.

<#registration-contacts>
    a solid:TypeRegistration ;
    solid:forClass schema:Person ;
    solid:instanceContainer </contacts/> .

<#registration-notes>
    a solid:TypeRegistration ;
    solid:forClass schema:TextObject ;
    solid:instance </notes/main.ttl> .

Two Registration Types

Property Meaning
solid:instanceContainer All instances of this type live in this container
solid:instance A single document containing instances of this type

Accessing the Solid Type Index

  1. Fetch the user's WebID profile
  2. Find solid:publicTypeIndex and/or solid:privateTypeIndex predicates
  3. Fetch the Solid Type Index document
  4. Query for solid:TypeRegistration entries with matching solid:forClass

Data Modelling Checklist

When modelling a new data type for Solid:

  • Check for existing SHACL shapes with types that meet your requirements
  • Check LOV for an existing vocabulary/class
  • Define a SHACL shape for the type
  • Add rdf:type statements to all instances
  • Register the type in the pod's Solid Type Index
  • Use stable, dereferenceable URIs for subjects
  • Add dcterms:created, dcterms:modified metadata
  • Document the shape in a public shapes repository
  • Validate data against the shape before writing to the pod

Resources

Resource URL
FAIR Principles https://www.go-fair.org/fair-principles/
Linked Open Vocabularies https://lov.linkeddata.es/
W3C SHACL Spec https://www.w3.org/TR/shacl/
W3C Data Shapes repo https://github.com/w3c/data-shapes
Solid type indexes https://solid.github.io/type-indexes/
Shapes resources https://github.com/w3c-cg/awesome-semantic-shapes
Solid Developer Documentation https://dev.solidproject.org/