Skip to main content

ActOfConnection (LinkedIn ontology module)

What it is​

  • A small set of Pydantic models representing LinkedIn-related ontology entities (Act of Connection, Person, Organization, etc.).
  • Each model can serialize itself (and linked entities) into an RDFLib Graph via a shared RDFEntity.rdf() method.
  • URIs are auto-generated per instance unless explicitly provided.

Public API​

Base​

  • class RDFEntity(pydantic.BaseModel)
    • set_namespace(namespace: str) -> None: sets the class-level namespace used for auto-generated URIs.
    • rdf(subject_uri: str | None = None, visited: set[str] | None = None) -> rdflib.Graph: emits RDF triples for the instance; recursively includes related entities while avoiding cycles.

Ontology entity models (all subclass RDFEntity)​

Each model has:

  • an auto-managed instance URI (_uri), plus ontology metadata (_class_uri, _property_uris, _object_properties)
  • Pydantic fields for data/object properties
  • rdf() inherited from RDFEntity

Public classes:

  • ActOfConnection
  • ISO8601UTCDateTime
  • Person
  • Organization
  • Location
  • ProfilePage
  • ConnectionsExportFile
  • CurrentJobPosition
  • CurrentOrganization
  • CurrentPublicURL
  • EmailAddress
  • ConnectionRole

Configuration/Dependencies​

  • Dependencies:
    • pydantic (models, validation)
    • rdflib (RDF graph creation)
  • Environment:
    • Default creator field uses os.environ.get("USER").
  • Namespace/URI behavior:
    • Default namespace for auto-generated URIs: http://ontology.naas.ai/abi/
    • Override via RDFEntity.set_namespace(...) (or per-subclass, since it’s a classmethod).

Usage​

from rdflib import URIRef
from naas_abi_marketplace.applications.linkedin.ontologies.modules.ActOfConnectionsOnLinkedIn import (
    ActOfConnection, Person, ISO8601UTCDateTime, RDFEntity
)

# Optional: set a custom namespace for generated instance URIs
RDFEntity.set_namespace("http://example.org/abi/")

alice = Person(label="Alice Doe", first_name="Alice", last_name="Doe")
ts = ISO8601UTCDateTime(label="2026-01-01T00:00:00Z")

act = ActOfConnection(
    label="Alice connected with someone",
    involves_agent=[alice],
    connected_at=ts,
    # object properties can also be URI strings/URIRefs
    occurs_in=[URIRef("http://example.org/location/nyc")]
)

g = act.rdf()
print(g.serialize(format="turtle"))

Caveats​

  • Many object-property fields default to the placeholder IRI string: http://ontology.naas.ai/abi/unknown.
  • created defaults are evaluated at import time (datetime.datetime.now() used as a direct default), so instances created later may share the same default timestamp unless explicitly set.
  • rdf() treats properties listed in _object_properties as object properties when values are str/URIRef; otherwise values are serialized as literals.