Skip to main content

Editor's Note: May 2024

Editors' note on the April 2024 issue of Reviews in Digital Humanities, guest edited by Kim Martin and Terhi Nurmikko-Fuller

Published onMay 28, 2024
Editor's Note: May 2024

Editors’ Note

Roopika Risam and Jennifer Guiliano

It’s been a really long month in a very long year. We’ve both been feeling a lot of disgust and dismay, from the militarization of college campuses; to state politics targeting LGBTQ+ communities, women, Indigenous nations, and DEI initiatives; to brutal suffering in Sudan, Gaza, Ukraine, and other communities around the world. If you’re feeling tired, disheartened, and deeply disappointed, we’re right there with you.

Running a monthly journal at the scale of Reviews in Digital Humanities, there’s a sense that the show must go on, even at the moments where “business as usual” feels unsuited to the moment — like this one. On one hand, having the privilege of engaging with the special issue editors, project directors, and reviewers, like those in this month’s issue, “Linked Data,” grounds us and gives us a sense of purpose through the service we’re providing to our colleagues. On the other hand, it can be a bit exhausting, especially when current events locally and globally are so demoralizing.

To give our incredibly hard-working team some well-deserved rest, Reviews will be on summer break from June 1-August 31, 2024. We have issue content ready to go and will continue publishing monthly throughout the summer. But we won’t begin processing new submissions that come in over the next three months until September. Our analysis of the patterns of reviewer responses has shown that following up on reviews over the summer takes significantly more labor than during the academic year, so we are conserving our team’s energy til September. We hope that you are able to do a bit of that yourself over the next few months.

In the meantime, we’re pleased to share the May 2024 issue of Reviews on Linked Data. We’re grateful to special issue editors Kim Martin and Terhi Nurmikko-Fuller for bringing their expertise on the topic to Reviews. Thank you, as well, to the project directors and the reviewers who have put great care into the content in this issue.

Guest Editors’ Note

Kim Martin and Terhi Nurmikko-Fuller

This special issue of Reviews in Digital Humanities focuses on projects utilizing the Linked Data information publication paradigm. Information aggregation can be achieved in many different ways — Linked Data is one of them. This approach — which involves making data available online in a machine-processable format, using existing Web technologies and architecture — utilizes a range of methodologies, theories, and standards to not just connect up disparate datasets but also to enable both human users and software agents to navigate complex graph structures. There is immense potential in two ways: in enrichment of knowledge by the aggregation of information from known and unknown sources alike and in the machine inference of implicit knowledge from explicitly declared facts. 

The quality and extent of Linked Data projects can be measured using the Five Star Open Data matrix. It rewards projects and dataset custodians for publishing data online, ideally in a structured format and utilizing non-proprietary formats: think of a spreadsheet published as a .CSV rather than an image of the spreadsheet published in a PDF. At the very heart of the Linked Data model are unique identifiers, which are assigned to data entities and the relationships between data entities.

A simple example might connect a person to their hometown. This example would be captured in three identifiers: one for the person, another for their hometown, and the third to denote the connection of belonging. These clusters of identifiers often occur in such clusters of three, a detail reflected in the nomenclature as we refer to these clusters as “triples.” Triples are a tangible, machine-readable representation of the RDF (or Resource Description Framework) abstract data model. RDF is used to represent data in a graph structure, which consists of any number of interconnected triples. In Linked Data projects we utilize existing Web architectures, and these identifiers are generally-speaking HTTP URIs (unique resource identifiers) — the type of thing we’re familiar with seeing in the browser address bar.  

The Linked Data paradigm enables information to be published in a machine-processable format. The resulting network of data entities and relationships is referred to as a “knowledge graph.” Because URIs in any given knowledge graph can point to any desired website, it is possible to aggregate information across the entirety of the Web. Linked Data is a method for bringing about the Semantic Web — a machine-processable, parallel version the Web we know (and love!). These URIs might themselves be part of other Linked Data projects, and the ultimate potential of this methodology is to aggregate all information online. Or, at least all information that has also been published using the RDF data model and HTTP URIs (i.e., as Linked Data).  

Knowledge graphs are self-referencing information representation structures. HTTP URIs capture both the type of data (for example, a person or a place), as well as the specific data instance that exemplifies that category (say, for example, Ada Lovelace as a specific example of a more generic idea of a person). In the nomenclature of Linked Data, these information categories are referred to as “classes” (relationships between classes are “properties”). A machine-readable document that tells the software and the human user which classes and properties are possible is known as an “ontology,” a term borrowed and appropriated from metaphysics. Ontologies can be used to show that different Linked Data projects have the same kind of information in them, even if the specific instances do not overlap. Whether alignments happen at schema-level or at the instance-level, Linked Data projects have almost endless potential for information aggregation and enrichment across the whole Web ecosystem.  

The vast potential of the Linked Data approach is illustrated by four example projects that range from proposographies to gazetteers, from mapping people to representing places. 

These projects showcase how tools such as ontologies, knowledge graphs, and triplestores can be developed and used to navigate complex datasets. The projects included in this issue cover topics that greatly range in time and space: 

  • WarSampo, WarVictimsSampo, and WarMemoirsSampo, three interconnected projects that create a rich and comprehensive picture of WWII data from Finland, created by the Semantic Computing Research Group (SeCo) at the Aalto University and University of Helsinki (Helsinki Centre for Digital Humanities HELDIG), led by Prof Eero Hyvönen, and reviewed by Michelle Meagher and Jana Smith Elford and by Sebastian Heath; 

  • Enslaved: Peoples of the Historical Slave Trade, which defines itself as a “discovery hub” for information about people involved in the historic slave trade, directed by Daryle Williams, Walter Hawthorne, and Dean Rehberger, and reviewed by Toby Burrows;

  • Pompei Artistic Landscape Project, an online resource that presents Pompeian artworks in their architectural contexts, directed by Eric Poehler and Sebastian Heath, and reviewed by Erin Canning; and

  • World Historical Gazetteer, an open access index of historical geographical places, directed by Ruth Mostern and Karl Grossner, and reviewed by Diane Jakacki.

The richness and diversity of these projects illustrate the flexibility, robustness, and the challenges presented by the Linked Data methodology. They serve as examples of the innovative research happening in this space in digital humanities, and we hope, serve to inspire a new generation of researchers to contribute to the ever-growing ecosystem of Linked Data. 

No comments here
Why not start the discussion?