Numismatic News: Roman and Byzantine departments updated in Mantis

Roman and Byzantine departments updated in Mantis

The next major wave of coins has been migrated from the ANS FileMaker database into the new curatorial database platform, CollectiveAccess. After several months of work in normalizing entities in OpenRefine, the Greek, Roman, and Byzantine departments (195,000 total objects) have been migrated into CollectiveAccess. The Byzantine and Roman coins have been published into the public database, Mantis, and the updated Greek department will be available online by tomorrow.

Altogether, more than 300,000 coins have been migrated into CollectiveAccess (roughly half the ANS collection), including the Islamic department (data cleaning undertaken by ANS contract cultural heritage data specialist Sami Norling) and the Medals and Decorations (which took me nearly six months to normalize nearly all of the artists, people depicted on coins, makers, issuers, etc.). Sami has subsequently finished cleaning the South Asian department, but it has not been uploaded into CollectiveAccess yet since I was busy with the Greek, Roman, and Byzantine objects.

This is a once-in-a-generation task, having inherited a database whose roots were in a bespoke MSDOS database from the 1980s. This database was imported into FileMaker in the late 1990s with no intermediate normalization or implementation of controlled vocabularies through relational database design.

When perusing the Greek, Roman, Byzantine, Medal, and Islamic departments in Mantis, the improvement in consistency and quality is readily apparent. This is especially true for the ancient and Byzantine coinage, since there is nearly complete coverage of numismatic concepts in Nomisma.org, including denominations, mints, regions, and people. Furthermore, nearly half of the Greek and Roman coins link to URIs published in online typologies, such as Online Coins of the Roman Empire and Hellenistic Royal Coinages. This had previously been the case in Mantis (which is one reason parts of the Roman and Greek collections were more consistent than others), but now we have linked tens of thousands of Roman Provincial Coins to URIs in RPC Online, thanks in large part to Jerome Mairat, who provided a concordance of ANS coins stored in their database. We did also link quite a few more that were not contained in the RPC Online database. In many cases, we were able to link to the issuer URI in RPC Online and make use of their preferred labels for individuals, thus greatly improving the human-readability of names in the issuer facet field.

Greek, Roman, and Byzantine findspots have been normalized to the best of my ability, and hoards have also been separated from the findspot field from FileMaker and normalized into controlled terms that can be used for query. The majority of ANS coins refer to hoards with IGCH URIs in CoinHoards.org, enhancing the map-based pages to show points on the query-able map for hoards as well as findspots, which were not geolocated in the previous iteration of data published in Mantis.

Mints and findspots of the Roman department, with finds from Britain to the Philippines

In addition to clear improvements in the controlled vocabulary terms in the search facets (authority, portrait, denomination, material, etc.), the indexing process pulls corporate entities and dynasties from Nomisma. Findspot and hoard are now search facets, making it possible to drill down specifically to coins from Boscoreale or other place. More than 140 Roman Republican coins now link to the Ancona 1 hoard in Kris Lockyear's Coin Hoards of Roman Republic, and points will appear on the map for these coins.

ANS 1944.100.238 from CHRR AN1

After the Greek department is published to Mantis, then I will reactivate the nightly updates. Unlike the FileMaker editing workflow, the modern APIs in CollectiveAccess enable us to run nightly updates of any object edited the previous workday, which significantly reduces the waiting time between the editing and publication of objects online. This has been a major logjam in the publication workflow since my arrival at the ANS in 2011.