WW1LOD: an application of CIDOC-CRM to World War 1 linked data

Eetu Mäkelä*, Juha Törnroos, Thea Lindquist, Eero Hyvönen

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

7 Citations (Scopus)


The CIDOC-CRM standard indicates that common events, actors, places and timeframes are important in linking together cultural material, and provides a framework for describing them. However, merely describing entities in this way in two datasets does not yet interlink them. To do that, the identities of instances still need to be either reconciled, or be based on a shared vocabulary. The WW1LOD dataset presented in this paper was created to facilitate both of these approaches for collections dealing with the First World War. For this purpose, the dataset includes events, places, agents, times, keywords, and themes related to the war, based on over ten different authoritative data sources from providers such as the Imperial War Museum. The content is harmonized into RDF, and published as a Linked Open Data service. While generally based on CIDOC-CRM, some modeling choices used also deviate from it where our experience dictated such. In the article, these deviations are discussed in the hope that they may serve as examples where CIDOC-CRM itself may warrant further examination. As a demonstration of use, the dataset and online service have been used to create a contextual reader application that is able to link together and pull in information related to WW1 from, e.g., 1914–1918 Online, Wikipedia, WW1 Discovery, Europeana and the Digital Public Library of America.

Original languageEnglish
Pages (from-to)333–343
Number of pages11
Issue number4
Publication statusPublished - Nov 2017
MoE publication typeA1 Journal article-refereed


  • Applying CIDOC-CRM
  • Data interlinking
  • Dataset
  • Historical data
  • Linked data
  • Modeling

Fingerprint Dive into the research topics of 'WW1LOD: an application of CIDOC-CRM to World War 1 linked data'. Together they form a unique fingerprint.

Cite this