Linked Conservation Data and Terminology: Modelling Workshop

User menu

Group photo of workshop participants. Photo by Gabriele Grigorjeva

By Athanasios Velios and Kristen St. John

On the 12th and 13th of September, the second workshop of the Linked Conservation Data project took place at the University of the Arts London. Funding from the Arts and Humanities Research Council in the UK allowed 30 participants to meet and discuss examples of conservation documentation and ways of abstracting them to produce unified data models for conservation. The workshop was divided into morning presentations and afternoon hands-on modelling sessions.


Linked Conservation Data discusses ways of sharing conservation data within and outside of the profession. The first workshop, which took place in Stanford, focussed on terminology: How can a conservator who is familiar with a term (e.g. endpapers) search records produced using a different but synonymous term (e.g. flyleaves)? The discussion led to the development of a workflow for publishing and matching terms from conservation vocabularies to enable this interchange during description and searching. We are preparing to test this workflow during the second phase of the project which, pending funding approval, will start in February 2020.

The subject of the second workshop was data structures; how can a conservator search conservation documentation from different institutions which, although they hold similar information, have been developed independently?

One way to think about this is by looking at the commonalities of documentation. It is possible to describe the broader questions that documentation may address:

· How something is made
· What the condition of something is
· What are the plans for treating something
· How something has been conserved

Within each of these sections we can capture data using similar abstract structures. Linked Data uses a structure of “triples” where there’s a subject, a predicate and an object. For the question of how something is made, one could describe the answer “the book cover is made of leather” as:

Cover (subject) --> made of (predicate) --> leather (object)

The definition of “leather” requires discussion around terminology as mentioned above. Describing predicates requires specifying how one item (the specific book cover) connects to another (the material of leather). To accomplish this consistently we utilize an ontology, which is a set of rules about how things are connected. An ontology widely understood to be useful for cultural heritage collections is the CIDOC-CRM ( In the example above, the predicate must be defined as “p45 consist of” which connects any object or component to its material. This straightforward structure can convey subtlety and complexity when several triples are used (and build upon each other) within a single treatment report.

During the workshop we looked at examples of conservation documentation and tried to specify these relationships according to the CIDOC-CRM. Participants provided sample records from their own institutions and then discussed these at the meeting. A list of the sample records is available on the LCD website (


The morning sessions included introductory presentations explaining the principles around integration of conservation records using a specific set of conceptual relationships between database entries provided by the CIDOC-CRM ontology. Examples of projects from the British Museum, the National Gallery, the Getty, the University of Oxford (EAMENA) and the Université de Cergy-Pontoise were presented. These presentations are being uploaded to the project website: (


During the afternoon hands-on sessions, participants were split into four groups broadly based on expertise. Each group was assigned a facilitator. Facilitators were chosen based on previous expertise with modelling. We were fortunate to have members of the CIDOC-CRM Special Interest Group participating in the workshop who took on facilitator roles including Dominic Oldman, George Bruseker and Robert Sanderson. Sample records shared by members of each group were then discussed inside the groups, and their structures were analysed.

During these exercises a number of particularly complex cases of documentation were identified as areas which require further discussion. For example, records about dimensions of objects are relatively straightforward to model and this was confirmed with many examples in the workshop. On the other hand records about non-existing features of objects (e.g. a book cover without decoration) are more difficult to describe given that we are recording something that does not exist. Additionally there was discussion in several groups considering how much modelling is necessary to offer information of value. Given the complexity of many treatment reports, at what point are there diminishing returns when modelling every single activity undertaken in a treatment? When is it sufficient for modelled data to point researchers to fuller records, and when is it important to include everything? We are hoping to continue with the discussion of conceptual modelling in conservation during the next phase of the project.


In January 2020 the project will hold its second webinar to summarise the work done over the first year of the project. The consortium is expanding, and we are looking forward to working with new partners. If you wish to be involved in the discussion or have interesting cases of conservation documentation to share, please post on IIC’s Community Platform ( under “Conservation Data”.


Dr Athanasios Velios is reader in documentation at the University of the Arts London as part of Ligatus, working on the documentation of conservation practice and modelling data for heritage conservation. He was trained as a conservator and has a PhD in computer applications to conservation. He was the webmaster of the International Institute for Conservation from 2009 until 2017.

Kristen St.John is head of conservation services for the Stanford Libraries. She was previously collections conservator at UCLA and special collections conservator for Rutgers. She has an MLIS and an advanced certificate in conservation from the University of Texas at Austin. Her interests include preservation education, the preservation and dissemination of conservation documentation, and historic bookbinding materials. 

(for the full article, see: Issue 75 "News in Conservation" December 2019)

Home Page Intro: 
How can a conservator search conservation documentation from different institutions which, although they hold similar information, have been developed independently? On the 12th and 13th of September, the second workshop of the Linked Conservation Data project took place at the University of the Arts London, to address this question.
Home Page Suppress Text: