Session 5

From Linked Open Data to RiC-O, LRM and CIDOC-CRM

Author

Stephen Hart

Published

13 October 2026

Modified

2 September 2026

Hart Taught by Stephen Hart.

In this session

The formal ontologies that archives, libraries and museums actually use:

RiC-O — Records in Contexts Ontology (archives)
LRM — IFLA Library Reference Model (libraries)
CIDOC-CRM — the reference model for cultural heritage (museums)

From your own model to a shared one

The Intro introduces RDF and Wikidata the same week. This session asks the next question: your model works for you — what does it take to make it work for everyone else?

Domain ontologies as negotiated agreements, and what the negotiation costs
Why three communities built three models, and where they converge
Mapping and alignment: what survives the translation and what does not

RiC-O — archives

Records in Contexts, and the move away from the single fonds hierarchy
Record resources, agents, activities, relations
Provenance as a first-class object rather than a note
What ISAD(G) could not express, and why archives went to a graph

LRM — libraries

Work, Expression, Manifestation, Item: the distinction, and why it is not pedantry
The FRBR inheritance, and what LRM consolidated
Modelling agents and nomen
Where LRM meets BIBFRAME and Wikidata

CIDOC-CRM — museums and cultural heritage

An event-centric model: things happen, and objects and people participate
E5 Event, E21 Person, E22 Human-Made Object, E52 Time-Span
Why “the vase was made in Athens around 500 BCE” becomes a production event with participants and a time-span
Extensions: CRMarchaeo, CRMtex, and the SDHSS ecosystem from the Intro’s reading

Exercise

Take one object or record from your own material and model it three times — once in RiC-O, once in LRM, once in CIDOC-CRM. Then compare: what did each one force you to state, and what did each one let you leave out?

In parallel in the Intro

Data modelling: Linked Open Data, RDF and Wikidata