This article is available at the URI http://dlib.nyu.edu/awdl/isaw/isaw-papers/7/elliott-heath-muccigrosso/ as part of the NYU Library's Ancient World Digital Library in partnership with the Institute for the Study of the Ancient World (ISAW). More information about ISAW Papers is available on the ISAW website.
Except where noted, ©2014 Tom Elliott, Sebastian Heath and John Muccigrosso; distributed under the terms of the Creative Commons Attribution License
This article can be downloaded as a single file
ISAW Papers 7.1 (2014)
Prologue and Introduction
Tom Elliott, Sebastian Heath and John Muccigrosso
Prologue
The articles published collectively as ISAW Papers 7 are the result of two meetings held at New York University's Institute for the Study of the Ancient World and at Drew University in early summer of 2012 and 2013 respectively. Both had the title "Linked Ancient World Data Institute" and came to be known by the acronym LAWDI. Organized by the present authors, the meetings were generously funded by the Office of Digital Humanities of the National Endowment of the Humanities. We are very grateful to the NEH and to our home institutions for their support of the events.
We would also like to thank all the LAWDI participants and all the contributors to this collection. At the LAWDI meetings, we strove for a combination of structured presentation and informal exchange, and here in the publication we intend to offer a forum for description of ongoing work. Different authors have made different use of the opportunity. A common thread is the willingness to make digital resources available on the public internet. But beyond that there is great diversity. Personal perspectives mix with brief reports and with technical discussions. Because all the contributions are informed more or less explicitly by the principles of Linked Open Data (LOD) it will be useful to review those here.
Introduction
Linked Open Data is a set of best practices for sharing digital resources. Among the foundational texts that helped define and promote LOD's principles is Tim Berners-Lee's "Linked Data" of 2006, which is online at http://www.w3.org/DesignIssues/LinkedData.html. Reaching further back into the history of the web, the same author's "Cool URIs Don't Change," online at http://www.w3.org/Provider/Style/URI, offers simple suggestions that are still useful today. Also influential is Sauerman et al. (2008) W3C Interest Group Note entitled "Cool URIs for the Semantic Web," for which see http://www.w3.org/TR/cooluris/. The ideas outlined in these three relatively brief discussions were described in much greater detail by Tom Heath and Christian Bizer in their 2011 book Linked Data: Evolving the Web into a Global Data Space, which is also available online at http://linkeddatabook.com/. Readers wanting to explore further approaches to LOD will be well served by "clicking through" on those links and also be investigating the resources at http://linkeddata.org.
Each session of LAWDI was only three days long so required considerable compression of technical details. We will do the same (and more) in this introduction. The one point that was stressed at LAWDI was stability, particularly of published URIs (and for discussion of the difference between URIs and URLs see Sauerman et al. [2008]). The fundamental practice of giving individual URIs to all discrete resources that a project published was energetically endorsed. Among the types of URIs often used as examples within the study of the ancient world were Pleiades URIs for ancient places (e.g. https://pleiades.stoa.org/places/59672) and the Worldcat catalog stood out as providing common identifiers with impact beyond any single discipline (e.g. http://www.worldcat.org/oclc/768969647). Beyond stability, the faculty and participants at LAWDI both stressed and searched for common vocabularies that can be used to encourage machine readability of the information available at the stable URIs that currently exist and that will become available. The Pelagios Project (Isaksen et al. 2014), which is aggregating disparate uses of Pleiades identifiers, has shown the utility of expressing links between resources in such a way that they can be automatically harvested. Nomisma.org has also made progress in this direction. Putting aside specific examples, the articles published here often express the desire for greater clarity on how to refer to people, places, and events, in addition to such abstract concepts as periods and typologies. That common practice is desirable was a clear outcome of the 2012 LAWDI session (Elliot, Heath and Muccigrosso 2012) and remained a topic of discussion in 2013.
One Technical Point...
Despite an avoidance of excessively technical presentations, LAWDI did not shy away from highlighting the “RDF Triple” as an important tool in ongoing efforts to encourage interoperability. We likewise think it probably that readers of the current essays will find it useful to be familiar with this fundamental concept in the publication of re-usable data on the public internet. Meaning that we offer here a very brief answer to the question, "What is a Triple?".
A triple is a statement of information in three parts. Using colloquial language, those three parts can be defined in the following way:
- The Subject: The thing being talked about.
- The Predicate: The category of statement one is making about the subject.
- The Object: The content of the statement that one is making about the subject.
For example, the English sentence "Athens is located in Attica" has the subject "Athens", the predicate "located in", and the object "Attica".
It is best practice in LOD to use URIs for all three elements of a triple so that the following three URIs are a machine readable version of the same sentence:
- https://pleiades.stoa.org/places/579885
- http://www.geonames.org/ontology#locatedIn
- http://www.geonames.org/6692632
Using the triple terminology, clicking on the "subject," or first URI in the list, will bring up the Pleiades record for Athens. Clicking on the predicate will bring up a human-readable introduction to the Geonames vocabulary. And clicking on the "object" will bring up the geonames.org record for Attica.
Showing that one simple English sentence can be transformed into a group of URIs is a prelude to the much grander assertion that triples can express many forms of information. An individual project's database might record object dimensions or site locations in a row and column oriented format that suits its particular needs. If that information is published on the Internet as triples, it is more likely to be usable by other consumers, whether human or automated. More complex series of triples can be linked together. A type of pottery is discovered in a building. That pottery has an origin and the building has a terminus post quem. All of this can represented as triples. If URIs are used, then those triples can refer to definitions distributed across the Internet. Data starts to be linked and analysis by third-parties is enabled. This is the vision to which LAWDI optimistically alluded.
Specific examples and discussions of triples appear often in this collection of essays, most directly in Almas et al., Heath, Kansa, and Romanello. The range of uses illustrated therein suggests the role that triples can play in providing a common format for representing diverse concepts published on the public Internet. It is optimistic to think that all digital data relevant to the ancient world will be published in this form, but it is also true that growing use of triples is leading to greater compatibility between the resources that are generating them.
Full presentation of the role of triples in LOD is beyond the scope of this introduction. We do encourage interested readers to see the useful discussion in Heath and Bizer's online book. Section "2.4.1 The RDF Data Model" is the most relevant part.
On This Collection
In closing we note that much diversity was on display at the LAWDI meetings and that the same is apparent in the scope of articles now being published as ISAW Papers 7. Our goal as editors is that the wide range of styles and approaches present in the face-to-face events will come through in LAWDI's written record. These are timely statements of ongoing work and we stress that they mostly relate to the state of affairs in late 2013. We hope that the issues this collection raises will be useful for others thinking about implementing digital resources. And we believe we speak for all LAWDI participants when stressing that the ideas articulated here are inspired by the desire to share well-conceived digital resources with all audiences willing to make use of them.
Works Cited
Berners-Lee, T. (1998). “Cool URIs Don't Change.” <http://www.w3.org/Provider/Style/URI>
Berners-Lee, T. (2006). “Linked Data.” <http://www.w3.org/DesignIssues/LinkedData.html>
Elliott, T., S. Heath and J. Muccigrosso (2012). Report on the Linked Ancient World Data Institute. Information Standards Quarterly, 24(Spring/Summer, 2/3): 43-45. < http://www.niso.org/publications/isq/2012/v24no2-3/elliott/>
Heath, T. and C. Bizer (2011). Linked Data: Evolving the Web into a Global Data Space <http://linkeddatabook.com/>
Sauerman L. and R. Cyganiak 2008. “Cool URIs for the Semantic Web.” <http://www.w3.org/TR/cooluris/>