A Semantic Web of Linked Data about Historic Bookbinding Archives

Sonja Heward-mills

Published in

OpenLink Virtuoso Weblog

5 min readOct 17, 2017

Situation Analysis

On the 28th of September, 2017, Athanasios Velios posted the following tweet:

https://twitter.com/a_velios/status/913401837163880449

And a few days later, @kidehen (Kingsley Uyi Idehen) posted the following, in response:

In these attention-challenged times, the intended goals of the tweet authors more than likely didn’t resonate with everyone interested the utility of a Semantic Web of Linked Data.

In this post, I flesh out the intentions behind these tweets in a manner that could also serve as a guide for understanding similar tweets from the past or in the future.

What is this about?

Athanasios Velios and the Ligatus team have embarked on a project to digitize documentation about bookbindings through the construction and publication of a Semantic Web of Linked Data.

Ligatus is a research centre in the University of the Arts London that focuses on projects related to historical libraries and archives.

A Semantic Web of Linked Data implies the following:

All entities (real and/or imagined) in this realm of discourse are identified by HTTP URIs (hyperlinks)
Each HTTP URI resolves to an RDF Document (content-types vary i.e., RDF-Turtle, JSON-LD, RDF-XML, and others manifest by way of negotiation between document-viewing clients and content-publishing servers)

In addition, this particular Semantic Web of Linked Data leverages Virtuoso (the same technology behind DBpedia — Linked Open Data Cloud seed and nucleus) which also adds the following to the functionality suite:

SPARQL Query Service Endpoint — HTTP-based Query Service for performing Data Manipulation operations against the published data
Faceted Browsing Service Endpoint — HTTP-based Service for exploring Entity Relationships that’s informed by Entity Relationship Type and Membership Count

Why is this important?

Data, Information, and Knowledge exposed by this documentation effort is highly discoverable and reusable, irrespective of data consumers being equipped with prior information about areas of interest. Basically, you have a rich database published to the Web that’s equipped with a high Serendipitous Discovery Quotient (SDQ).

How do I exploit this power?

Assuming you are equipped with nothing more that a link to the Ligatus SPARQL Endpoint, here’s how the combined prowess of follow-your-nose (FYN) exploration and SDQ manifest:

[1] Go to http://data.ligatus.org.uk/sparql — and then select “HTML (Faceted Browsing Links)” from the “Results Format” menu.

Selecting Query Results Option for Faceted Browsing based on Query Results

Alternatively, you can edit the query as depicted below, such that Entity Identifiers (HTTP URIs) and Entity Labels appear in the Query Results.

Modified Default Query that adds Entity Labels and Concept Type scoping

[2] Click on the “Run Query” button which leads to a Tabular Query Results Page — where each HTTP URI in the “entity” column is an Entity Identifier and each chunk of text in the “label” column is an Entity Label .

[3] Aided by Entity Identifiers and Entity Labels, click on an Entity Identifier that grabs your interest — which results in an Entity Description Page that lists all the attributes of the entity selected

For instance, if the Concept Scheme (Taxonomy) created by these domain experts is of interest to you, simply click on the hyperlinked text “Language of Bindings” which will take you to a page that unveils a description of the Language of Bindings Taxonomy produced by this project.

From this point on, you can continue to explore or share information of interest by clicking on the “permalink” button and sharing the link presented with others via email, tweet, discussion thread (internal or external), etc.

Conclusion

Fusing “Semantics” (i.e., Entity Relationship Type Semantics) and “Web” is all about using the infrastructure already provided by the Web (Entity URIs, Document URLs, HTTP) to bring the logic behind Relationship Type Semantics to life in a manner that delivers a much richer experience for knowledge production, consumption, and reuse.

In this example, domain experts from the Ligatus Project have successfully published knowledge about bookbinding and archiving in a form that’s highly discoverable and flexible to consume; i.e., there are no preordained RDF document content-types — HTML is used as the default due to its common support across all browsers.

Finally, by leveraging Virtuoso as the deployment platform (providing Linked Data, Faceted Browsing, and SPARQL Query Services), this powerful point of presence provided by the Ligatus Project is immediately available to a wide range of pre-existing tools including those that support any of the following open standards:

HTTP
ODBC
JDBC
ADO.NET
OLE DB

Most importantly, all of this is achieved without compromising performance or security.