Making amateur archaeological collections accessible via Linked Open Data

Portable Antiquities of the Netherlands renders data useful to both the scientific community and the public.

10/17/2019 | 4:28 PM

FotoStijnHeerenKleinIn 2018, VU archaeologist Stijn Heeren and his team won the Dutch Data Prize for their project Portable Antiquities of the Netherlands (PAN). 'The jury really valued the way we're sharing the data through the PAN website, archiving it with DANS-Easy and structuring it via Linked Open Data. They also praised our use of citizen science', Heeren explains enthusiastically. 

Documenting amateur collectors’ finds
'In the PAN project, we have four years (2016-2020) to document the archaeological finds of amateur collectors, particularly those made by metal detectorists. Metal detectors first became widely available in the 1970s, at which point they began to be used by hobbyists. While the collections of such detectorists are of great scientific value, they have largely been excluded from archaeological research until now. Through PAN, we're making them available for scientific research as well as for other amateurs.’

Most important, Heeren emphasises, is now to document the amateur collections as quickly as possible. 'The reason is that the collectors are getting on in age and it is crucial that they tell us where the finds were made. Once our project is wrapped up, the Netherlands Cultural Heritage Agency will take care of documentation for any new amateur finds.'

Providing access for both the scientific community and the public
The objective of the PAN project is to document the amateur finds and publish them online in order to make the information on the objects as well as the locations where they were found available to the scientific community, cultural heritage research, museums and any other interested parties.

Access through Linked Open Data
'Since the new Heritage Act entered into force in 2016, amateurs have been required to report anything they find. We visit the people at their homes to take pictures of their finds – things such as fibulae, keys, coins or buckles – and ask them where they found them. Next, we describe and publish the objects online via Linked Open Data. What makes PAN so innovative is that we work with standardised object descriptions. These descriptions consist of links to existing data sets and thesauri. With regard to terminology and archaeological periods, we rely on existing thesauri whenever possible. When we do introduce new terms, we indicate how they relate to other, similar concepts.'

Advantages of Linked Open Data
'Linked Open Data (LOD) make it possible to exchange information with other systems, such as external databases, much more efficiently. Entries in a traditional database and old-fashioned hyperlinks refer to static data. When they are adopted by a third party and we then update our information, the outdated copy remains in the other location. By working with live links rather than copying data, the information is automatically updated everywhere when adjustments are made to the source.'

56,000 objects with 3,500 reference types
'To date, our data set of finds excavated by amateurs contains 56,000 objects with 3,500 reference types. There is a great deal of variety in the types of finds. PAN is the first project that provides digital access to archaeological finds by type. What's more, new objects – which were previously unknown in the professional literature – are now being discovered. At least 80 never-before described garment pins (fibulae) have been found!' Amateur collectors do their part in facilitating access to the objects as well, Heeren says. 'The metal detectorist community has a real wealth of knowledge. For instance, while hardly any professional literature on medieval keys exists, there is lots and lots of collectors’ literature. These collectors are now drafting the reference types of medieval keys for us.' StijnHeerenSoloKlein

Not everyone can see and access all information
'Every time we document a find, it is published on the PAN website. However, not everyone can see and access all the information', Heeren stresses. 'Members of the general public who visit the website will see only the municipality in which the object was found, as the finders prefer not to let everyone know the exact location. The precise site where the object was found is on record in our secure database, however. Archaeologists can request our permission to use the data and database or get an export of the data if they like. An increasing number of archaeologists from the Netherlands and abroad are taking advantage of this. Finally, the thesaurus data are accessible to the public via the PAN website, while they are also available via servers of archaeological and other institutions.'

Long-term archiving and good findability thanks to DANS-Easy
'To store the data, we use the servers at the University of Groningen's Geodienst. We also deposit a back-up of all our data with DANS-Easy. They make sure the data stay accessible in the long term, regardless of how the software will change in the future. Storage with DANS also means the data are easy to find; there, it is readily accessible to European servers such as Ariadne as well as Europeana in the future.'

Improving the searchability and accessibility of the PAN website
The winner of the Dutch Data Prize will receive €5,000 to make their data set even more accessible. What does PAN intend to do with the money? Heeren: 'We plan on using it to improve the searchability and accessibility of the PAN website for the general public. We want them to be capable of viewing and consulting the PAN data set even more effectively.'

About Portable Antiquities of the Netherlands
PAN comprises a staff of seven registrars who meet with collectors at their homes or in museums so as to document the collections, various finds specialists who provide the scientific determinations, and the project managers.
https://portable-antiquities.nl

See also:
‘The value of metal finds in archaeology – Why PAN’, by VU archaeologist and PAN programme coordinator Stijn Heeren.

About the Dutch Data prize
The Dutch Data Prize is awarded every two years to a researcher or research group who makes extra contributions to science by making research data available for additional or new research. Its next edition will be in the autumn of 2020; it will be possible to nominate yourself or others from the start of 2020. Visit https://researchdata.nl/diensten/dataprijs/ to learn the criteria for nomination.