Using Linked Open Data to Navigate the Past: An Experiment in Teaching Archaeology

Introduction
While there is increasing research about scholarly publication workflows through Linked Open Data in Archaeology (even from the viewpoint of non-specialists [Seifried, 2014; May et al., 2012]), little has been done to take advantage of its potential for teaching. However, LOD is going to be the main method of publication and outreach in the discipline: therefore, it is becoming increasingly important for students and teachers to become familiar with its structure.
Linked Open Data is a powerful tool for navigating through the complexity of the inherently multifaceted reality of archaeological sites, which results from the intersections of space, materiality, language, visual culture, history, text, and so on. However, LOD also poses the challenge of how to manage such complexity in a meaningful way. In this paper, we report on an experimental project developed during a Classical Archaeology course in 2018, during which we researched four different Graeco-Roman sites, with the goal of reconstructing the main aspects of their material history through exclusively LOD-based resources.

The experiment
We explored four sites: Ostia Antica, Herculaneum, Alexandria, and Eleusis. Because the designed ground of the exploration was geographical, we started by using resources that focused on location as the main interlinking structure. We started with

Peripleo
(Simon et al., 2016), the LOD-based search engine of Pelagios, a project whose main purpose is to connect together several partnered datasets, by using places as the common ground (Isaksen et al., 2014).

Preliminary choosing the sites to explore enabled to establish a certain hierarchy in the information: first, we gathered the resulting links to gazetteers (especially
Pleiades [Bagnall et al., 2018]), to collect information about the site, including relevant geographical connections and attested chronologies. We then used Peripleo’s internal vocabulary to isolate specific periods of use of our sites, and classified the resulting links according to the database of provenance.

Initially, we managed the richness of the results by focusing on specific typologies of information: so, we harvested links to

Fasti Online
(2004-) to collect data about excavations in our sites; we looked for artifacts through a large variety of databases, including

Ariadne
(2012),

Arachne
,

WikiData
,

Vici
(Voorburg, 2014), the

Perseus
(Crane, 2018), and

Flickr
; we used these resources to find pictures and basic information, and as a starting point towards other datasets from museums and archives.

Coinages and hoards were a prominent feature, which was explored through the resources of the American Numismatics Society (2016),

Nomisma
, and the

Münzkabinett Online Catalogue
of Berlin. Textual sources were mainly collected through

ToposText
, which was privileged for its focus on locations. Inscriptions were found through the

EAGLE Network
and the

University of Graz Online Portal
.

From each of these resources, we were further directed in our exploration according to the emerging patterns in our findings: for example, the information collected through

Arachne
was extremely useful to access bibliographical data on religious artifacts in Eleusis; we used

Europeana
to further research Herculaneum, and from there we accessed the Bodleian Library collection of the Herculaneum Papyri and the related information about their context of finding, the Villa of Papyri, on
Pleiades (Bagnall et al., 2018). Researching Coinages of rulers in Alexandria provided a significant amount of information about the Roman emperors that controlled the city, through links to biographical databases (e.g.

VIAF
). EAGLE enabled us to even access
not LOD-based resources, such as the

Epigraphic Database of Rome
, to collect alternative transcriptions.

Results
We created object cards, specifying the essential information, the online identifier, and the further connections to other findings or contexts. Collectively, we were able to assemble approximately 50 cards, which, depending on the specificities of the site, included excavations, inscriptions, coinage, textual sources, monuments, mosaics, houses, papyri, sculptures (in bronze, marble, terracotta), pottery, and even furniture remains. We then wrote reports about each site, focusing on the best attested chronological periods (Herculaneum) or on prominent aspects of material history: in the case of Alexandria, for example, we propose an exploration of material evidence documenting cultural syncretism in the Graeco-Roman era.

Pedagogical and scholarly outcomes

Discoverability: through Linked Open Data, students are able to discover a variety of archaeological information with scientific reliability. Provided that the tools used are adequate to the purpose, this kind of exploration has a strong potential for non-experts, which is extended to the discovery of less known museums and archives around the world.
Complexity: Linked Open Data enables to navigate through an immense set of records, discovering unexpected and continuously new types of findings. This contributes to create a rich picture of the complexity of an archaeological site, which is the result of several interconnected pieces of evidence.
Contextualization: LOD makes it possible to explore contextualized information. Students are not only able to look for a specific artifact, but they can dig into the data that provide information on the context where it was found, the people and organizations related to it, the events connected to its chronology (Meadows and Gruber, 2014).
Effective scholarly digital research: students can perform complex research across a variety of different resources, and learned how to navigate through the information of specialized datasets with very different structures, vocabularies and functions.

Limitations

There is still too much volatility in the adoption of shared vocabularies across LOD resources: this limits the range of searchable items, specifically types of archaeological findings. This is a well-known problem in the world of Archaeological Linked Open data, now further reinforced by user experience (Bechhofer et al., 2013; May et al., 2015).

While there has been much investment in LOD infrastructures, the actual availability of data and metadata is still questionable for some archaeological sites. Whereas sites like Ostia and Herculaneum display a considerable amount of semantically diversified and interlinked material, places like Alexandria are more of a challenge in researching accurate and rich information, for the scarce availability of digitized archaeological records.

Bibliography

American Numismatics Society
(2016),

http://numismatics.org/

(accessed November 12, 2018).

Ariadne Infrastructure
(2012),

http://www.ariadne-infrastructure.eu/

(accessed November 12, 2018).

Bagnall, R., et al. (eds.),
Pleiades: A Gazetteer of Past Places
,

http://pleiades.stoa.org/

(accessed November 12, 2018).

Bechhofer, S., Buchan, I., De Roure, D., Missier, P., et al. (2013), Why Linked Data is not Enough for Scientists.
Future Generation Computer Systems
, 29(2): 599-611.

Crane, G. R. (ed.),
Perseus Digital Library. Artifact Browser
,

http://www.perseus.tufts.edu/hopper/artifactBrowser

(accessed November 12, 2018).

EAGLE Network
,

https://www.eagle-network.eu/

(accessed November 12, 2018).

EDR: Epigraphic Database Roma
,

http://www.edr-edr.it/

(accessed November 12, 2018).

Europeana
,

https://www.europeana.eu/

(accessed November 12, 2018).

iDAI
.
objects - Arachne
,

https://arachne.dainst.org/

(accessed November 12, 2018).

Fasti Online
(2004-2018),

http://www.fastionline.org/

(accessed November 12, 2018).

Flickr
,

https://www.flickr.com/

(accessed November 12, 2018).

Isaksen, L., Simon, R., Barker, E. T. E., de Soto Canamares, P. (2014). Pelagios and the emerging graph of ancient world data.
WebSci ’14: Proceedings of the 2014 ACM conference on Web science
, ACM, pp. 197–201.

May, K., Binding, C., Tudhope, D., Jeffrey, S. (2012). Semantic Technologies Enhancing Links and Linked Data for Archaeological Resources.
Revive the Past. Computer Applications and Quantitative Methods in Archaeology (CAA). Proceedings of the 39th International Conference, Beijing, April 12-16.
Amsterdam: Pallas Publications, pp. 261-272.

May, K., Binding, C., Tudhope, D. (2015). Barriers and Opportunities for Linked Open Data Use in Archaeology and Cultural Heritage.
Archäologische Informationen
, 38: 173-184.

Meadows, A., Gruber, E. (2014). Coinage and Numismatic Methods. A Case Study for a Linking Discipline.
ISAW papers
7: 15 (

http://dlib.nyu.edu/awdl/isaw/isaw-papers/7/meadows-gruber/

).

Münzkabinett - Staatliche Museen zu Berlin
,

https://ikmk.smb.museum/

(accessed November 12, 2018).

Nomisma,

http://nomisma.org/

(accessed November 12, 2018).

Peripleo, an initiative of Pelagios Commons
,

https://peripleo.pelagios.org/

(accessed November 12, 2018).

Seifried, R. M. (2014). Linked Open Data for the Uninitiated.
ISAW Papers
7: 26 (

http://dlib.nyu.edu/awdl/isaw/isaw-papers/7/seifried/

).

Simon, R., Isaksen, L., Barker, E. T. E., de Soto Cañamares, P. (2016). Peripleo: a Tool for Exploring Heterogeneous Data through the Dimensions of Space and Time.
Code4Lib Joyrnal
31: 2016-01-28 (

http://journal.code4lib.org/articles/11144

).

The Bodleian Libraries and University of Oxford,
Digital Bodleian
,

https://digital.bodleian.ox.ac.uk/

(accessed November 12, 2018).

ToposText 2.0
, © Aikaterini Laskaridis Foundation,

https://topostext.org/

(accessed November 12, 2018).

University of Graz, Institut für Archäologie - Zentrum für Informationsmodellierung - Austrian Center for Digital Humanities.
Online Portal der Archäologischen Sammlungen

http://gams.uni-graz.at/

(accessed November 12, 2018).

VIAF
-
Virtual International Authority File,

https://viaf.org/

(accessed November 12, 2018).

Voorburg, R. (2014),
Vici.org: Archaeological Atlas of Antiquity
,

https://vici.org

(accessed November 12, 2018).

WikiData,
https://www.wikidata.org (accessed November 12, 2018).