Linked Past 6

Page 1

Linked Pasts 6 poster hall: Poster board 1 John Bradley, DPRR and POMS as RDF: Challenges and Promises Samuel Clark & Shai Gordin, MAPA: Mapping Ancient Uruk by Combining Gazetteer Analysis and Remote Sensing Elizabeth Fentress, North African Heritage Archives Network Poster Board 2 Nitzan Gado & Sinai Rusinek, Feeding a Gazetteer with Word2Vec Sara Graveleau, Clarissa Stincone et al., Linking Authors and Works in the Dictionnaire Universel by Basnage de Beauval Poster Board 3 Karl Grossner & Ruth Mostern, Linked Traces: Connecting places via historical events, people, objects and concepts Ethan Gruber, Renee Gondek & Tyler Jo Smith, Poster Board 4 Piraye Hacıgüzeller, C. Verbruggen & P. de Potter, Collaborative Linked Data Creation Platforms for the Arts and Humanities Alexander Ilin-Tomich, Codifying (un)certainties in an ontology for a prosopography of the Egyptian Middle Kingdom Poster Board 5 Rebecca Kahn, Valeria Vitale et al., From Linking Pasts to Linking People: a Pelagios (virtual) Marketplace Lance Martin, Linking Entity Annotation in the Coptic Scriptorium

Poster Board 6 Christine Plumejeaud-Perreau, Mélissa Mimouni et al., PORTIC: a gazetteer for French ports (1749–1815) Jessie Pons & Frederik Elwert, DiGA and Beyond: Digitization and Linked Open Data for Gandhāran Art History Poster Board 7 Pere Pau Ripollès, Manuel Gozalbes et al., Setting up a comprehensive digital catalogue of the coins of Ancient Spain Matteo Romanello, Distributed Text Services (DTS): A Communitydriven Text Interoperability API Poster board 8 Donald Sturgeon, Crowdsourcing Chinese history: distributed transcription, annotation, and datafication Florian Thiery, Allard Mees & Dennis Gottwald, Linked Open Samian Ware: Transformation and modelling of vagueness and uncertainty Poster board 9 Andreas Wagner & Annemieke Romein, Entangled Ordinances: Linked Open Data for Legal History Annemieke Romein & Vany Susanto, Early Modern Ordinances in the Dutch Republic and Dutch East India Company’s territories Fernanda Olival, Renata Vieira et al., PM_1758: From Handwritten Records to its Links

DPRR AND POMS AS RDF: CHALLENGES AND PROMISE John Bradley (KCL, Digital Humanities, King’s Digital Lab) Prosopographies are, by their very nature, “linked data” about the past. Two of the many data prosopographies created at KCL’s DDH and KDL have been expressed recently as RDF with RDF servers. By making them we were exploring how RDF (described by Tim Berners-Lee as the “language of linked data”) works to make material of this kind available to a research community. • SPARQL querying seems to provide the most benefit… • All 4 kinds of visualisations shown to the right draw their data from these RDF servers in this way. What kinds of things can be done by a researcher with RDF & SPARQL that cannot be done with the UXoriented browser front ends? • Can historians actually exploit the material expressed in these technologically-heavy technologies? There’s no room for much information on an A4 size poster like this. For more info, see: • A recent (vol 14 No 2) DHQ article A Prosopography as Linked Open Data: Some Implications from DPRR; 000475.html • A talk given at the Digital Classicist London Seminar, 26 June, 2020;

Questions? Email them to

People of Medieval Scotland (PoMS) • The RDF server is at • See the documentation about it at • PoMS contains data suitable for SNA work. • PoMS contains geographic data that can be mapped. The map (right) shows land exchanged between women in PoMS. • The data was extracted using a SPARQL query, and mapped through a small Python script. • See • See the tutorial about SPARQL on PoMS at

Digital Prosopography of the Roman Republic (DPRR) • The RDF server is at • See the documentation about it at • Prof Chris Johanson (UCLA: Classics) and his students have done original work with DPRR RDF entirely independently of KCL (samples to the left). See: • “Shape of Roman history”: • “Funerals on the Rostra”. In UCLA RomeLab project.

The North African Heritage Archives Network NAHAN ( is a new international project aimed at creating a platform for documents from archaeological archives held in European and North African institutions. Initiated in February 2016, the partnership is a loose network, linked initially by a Memorandum of Understanding. The project functions under the international ‘umbrella‘ of ICCROM-Athar. A platform is now being created jointly by the CNRS laboratories of the Centre Camille Jullian (CCJ) and Archéologie & Philologie d’Orient et d’Occident (AOrOc) . The platform will be an open-access resource, and will both harvest data from other sites where documents are already online, and provide sustainable space to hold digitized documents. It would also allow the posting of catalogue entries rather than documents, in the case of an institution holding documents catalogued but not yet digitized. With this discussion we aim to open the project up to others interested in online access to archives, in order to discuss the problems involved, both in terms of the human relationships involved (the diplomatic side) and the technical ones. What are the mechanisms for gathering up the available data and making it discoverable? What are the possibilities of finding financing for cataloguing and digitization? What is the best way to align different resources? What is the relevance to EU Projects like ARIADNE+? Charlotte Roueché, Philipp von Rummel, Christophe Goddard, Stéphanie Satre, and Sebastian Stride will take part in the discussion, moderated by Elizabeth Fentress. All interventions are welcome!

Feeding a Gazetteer with Word2Vec KIMA, which started thanks to two Pelagios grants, is a historical attestation-based gazetteer for Jewish Languages in the Hebrew Script. Our Historical Hebrew Newspaper Corpus contains over 9 million words in over 25 thousand textual records published in the Hebrw newspaper “HaZfirah� during the years 1862-1888.

Nitzan Gado & Sinai Rusinek

Where do we get the variant toponyms to populate a gazetteer? Next to digitizing print Gazetteers, querying existing data sources such as Wikidada and Geonames for the specific domain in which one is interested, and mining them manually or with the help of Named entity recognition software, we turn to another method: starting with known toponyms in a corpus and leveraging them to detect those that are hitherto unknown to us, using word embeddings:

Initial set: 183 variant names for 141 places.

Word embeddings were created using Gensim with skip gram. The window size was 5 and minimal count 2.

In the embedding space, 73% of the 2960 tokens closest to toponyms from our initial set (blue) were also toponyms, which often also exhibit some interesting similarities.

Of the 893 unique toponyms found, 332 were known toponyms in our Gazetteer, 561 were new variants of places from our gazetteer, and we even found over a dozen names of places which were not yet included in the Gazetteer. The work began as a course assignment under Dr. Yuval Feinstein, the Hebrew University of Jerusalem The corpus was created at the OMILab historical press project, in collaboration with Historical Jewish Press .

Contact us:

One dictionary, several editions Linking Authors and Works in the Dictionnaire Universel by Basnage de Beauval (1701) Sara GRAVELEAU, Clarissa STINCONE, Geoffrey WILLIAMS, Ioana GALLERON

General workflow

PDF file


Types of Named Entities (NE)

TXT/ double PDF


XML file

XSLT + enrichment

Final XML + Web interface

a) b) c) d) e) f)

Furetière, 1690

Basnage, 1701

Brutel, 1725

AUTHORS (full names; last names; abbreviations) WORKS (full names; abbreviations) PLACES HISTORICAL PERSONALITIES ORGANISATIONS SPECIFIC EVENTS

Manual annotation and training

Named Entities Recognition and Disambiguation (NERD) Workflow

SEM (SegmenteurÉtiqueteur Markovien)

Preannotation and manual correction of .txt files

Training (CamemBERT/ OSCAR)

Injection of NE in XML file

addition of ISNI codes, links to Wikipedia

extraction tool

Linked Traces

Connecting places via historical events, people, objects and concepts

In Theory

The geography of { X }

Gazetteers, databases of named places, describe relationships between people and the world. After all, place names reflect linguistic, cultural, and political authority. When gazetteers are historical, they also describe how the power to name places changed hands. However, other than simple relationships of containment, gazetteers cannot describe connections between places.

Therefore, we have developed traces to assert relationships between places and non-spatial phenomena. Traces attest that given places were waypoints on journeys, centers of artistic movements, or residences of individuals. Since journeys, movements, and lifespans entrain multiple places, traces are a solution for how to illustrate complex and heterogeneous human geographies.

Theme {X} Project 1

Theme {X} Project 2

In Practice Peripleo

The Peripleo search engine developed by the Pelagios project indexes trace annotations in addition to place records. A record displayed for a given place may therefore include images of historical objects for which the place was a findspot - for example coins, hoards, and inscriptions, and links to contributors' records of the objects.


traces are historical things of interest, both material and non-material objects



Karl Grossner Ruth Mostern Susan Grunewald Ali Straub

concepts romanticism

World Historical Gazetteer trace data are web-published descriptions or depictions of a trace or traces http://.../...






trace annotations associate trace data URIs with places related to them (and how & when) Web Annotation records

Target (trace data)




The World Historical Gazetteer project has incorporated an index of trace annotations into its platform as well, and has begun developing trace annotation data for historical events and processes, e.g. journeys and expeditions, lifepaths, and commodity trade.

Trace annotations of the lifepath of Gautama Buddha and a journey of 7c. Buddhist monk Xuanzhang illustrating their intersection 1000 years apart.

A trace of colonialism and empire: annotations of a dataset about Hernan Cortes' Conquest of the Aztec Empire with its relevant locations, including waypoints and battles.

Body/Bodies (related places) http://.../...

http://.../... http://.../...

what else ??


● ● ● ●

Codifying (un)certainties in an ontology for a prosopography of the Egyptian Middle Kingdom

Same person Stela Ashmolean AN1971.5 wab-priest Auibhor


project Persons and Names of the Middle Kingdom (PNM)

or two namesakes

Stela Hildesheim 4589 wab-priest Auibhor

The dilemma is unsolvable, but this ambiguity is lost in formal ontologies Common approach (Factoid Prosopography Ontology, LAWD/SNAP:DRGN) linked through Assertion (factoid) referenced to Source or

Person Has bonds with other persons, life facts (mediated by sources in FPO)


linked though Attestation

Proposed approach (PNM ontology)

> hieroglyphic lexicon of personal names, index of titles, prosopography > timeframe: 2055–1550 BC > ~9000 sources > ~35000 name attestations > ~5000 names, ~15000 spellings > ~2500 dossiers with >1 source > a one-man project; research and coding in the same hands > planned completion Aug 2021


> relational db MS Access, MySQL doi:10.5281/zenodo.1411391 > PHP frontend Is carried by an object, Has bonds with other has a status Has bonds with other with physical data, conjectural persons, (accepted/rejected/weak) persons in the same museum inv., archaeo- > interim ontology, relying on conjectural life facts and reasoning source, life facts as known logical context CIDOC CRM, CRMEH, Ontolex (e.g., same father, same title) from the source > own R2RML mappings factors of uncertainty why codify uncertainty links & examples > few biographical and genealogical > of all identifications proposed in past > > db2triples (open-source tool to convert MySQL to triples based details in inscriptions and papyri research, only reliable ones are suited > SPARQL endpoint and web console on R2RML mappings) > common names (20 most popular for analyses > Jena Fuseki as triplestore and names account for 19% of all known > data with different level of reliability > sample person graph (Auibhor) SPARQL-endpoint (in work) people in Middle Kingdom Egypt) suit different research objectives Conjectural person

← refers to ←

Identity statement

← has identity statement ←

Person in a source

→ is re

ferred to by →

Alexander Ilin-Tomich Johannes Gutenberg University Mainz

Source text

From Linking Pasts to Linking People: a Pelagios (virtual) Marketplace What is Pelagios?

Aim of this session

The Pelagios Network connects researchers, scientists and curators to link and explore the history of places. The activities of Pelagios are:

Do you use linked data to interconnect historical sources? Is annotating places of the past a passion? Would you love to geo-visualise your dataset of people, places and sources but don’t know where to start? Then this session may be for you. After 18 months of existence, the Pelagios Network has grown into a diverse group of projects and individuals with shared interests in linked data, historical sources and representations of knowledge. While there is a wealth of activity throughout the network, interconnecting people with related activities and interests remains a little fragmented. To help us boost these interconnections, this Pelagios Marketplace session proposes a (virtual) conference space during Linked Pasts for newcomers to the network to learn about existing tools and activities, participants to present their activity, ask for specific support in terms of skills or experiences, share best practices or dreams for new functionalities for their favourite tools. We envision inviting participants to share specific requests, as well as general observations, in order to take advantage of the broad range of people and projects that constitute the Linked Pasts and Pelagios Network communities.

World Historical Gazetteer Peripleo and Linked Traces Linked Traces

● World Historical Gazetteer:


Traces are historical things of interest, which may be material or non-material, e.g. events, people, archaeological objects, artistic works, and artistic or philosophical movements. Trace data are web-published descriptions or depictions of traces, in the form of texts, datasets, images, or some combination of these. Trace annotations associate URIs for one or more trace with URIs for one or more records of places relevant to a trace or traces.

Peripleo is a linked data search engine search engine developed by the Pelagios project. It indexes trace annotations in addition to place records, and allows users to view aggregated linked data collections in a map-based interface. Peripleo was a pilot project that is not currently maintained, but the concept of linking records of historical entities with records for places relevant to them remains a compelling one. The World Historical Gazetteer project has incorporated an index of trace annotations and has begun developing trace annotation data for events, e.g. journeys, lifepaths, and expeditions of conquest.

The World Historical Gazetteer web platform (WHG) aggregates attestations of places drawn from historical sources by researchers studying the past within and across many disciplines in the humanities and social sciences. Registered users can: ● upload place datasets into a private workspace ● augment them via reconciliation against Getty TGN, Wikidata, and the WHG index itself, then download results ● publish them as Linked Open Data by flagging them as 'public', accessible via an API ● contribute them for accessioning to the WHG index, making them accessible in the site search engine Although WHG v1 has considerable functionality, it is at an early stage of development, and the project team is seeking feedback on its features, usability, and usefulness, as well as data contributions. A Site Guide explains existing functionality in some detail, and WHG demo and discussion sessions will be part of the “Linking Geo-data Through Test and Play” activity during Linked Pasts 6, on 10 and 11 Dec, at 15:00 GMT .


Programme 1. 2.

Recogito is an online platform for collaborative document annotation. It is maintained by Pelagios, a Digital Humanities initiative aiming to foster better linkages between online resources documenting the past. Recogito provides a personal workspace where you can upload, collect and organize your source materials texts, images and tabular data - and collaborate in their annotation and interpretation. Recogito helps you to make your work more visible on the Web more easily, and to expose the results of your research as Open Data


4. 5. 6.

Kahn, R., Vitalie, V., Grossner, K., Barker, E., and Chambers, S. (2020) From Linking Pasts to Linking People - a (virtual) Pelagios Marketplace - a poster and virtual marketplace session at: Linked Pasts 6, University of London and British Library, December 2-16, 2020

Welcome Ice breaker (a fun game to gauge the level of knowledge about the Pelagios network from the participants) Lightning talks of 5 minutes each about 5 Pelagios tools / activities (e.g. Peripleo & Linked Traces; World Historical Gazetteer; Recogito, WebMaps-T… ) 5 lightning talks from researchers / projects Brainstorm session 25 minutes to exchange ideas / priorities (please feel free to add your ideas to our Google doc) 5 minutes wrap up of next steps

Some of our network members:



Coptic Scriptorium


The Pelagios Network: connecting researchers, scientists and curators to link and explore the history of places @PelagiosNetwork | | Background image courtesy of Karl Grossner (

This work is licensed under a Creative Commons Attribution 4.0 International Licence

Lance Martin Digital Humanities Specialist, Coptic SCRIPTROIUM

Visualizing the data Entity Types have

Linking Named Entities What is Entity Annotation Entities are referring expressions (noun phrases). Entity annotation labels entities according to both entity types and named/non-named entities.

Coptic SCRIPTORIUM links mentions of named entities to their corresponding Wikipedia articles, i.e. Wikification.


‘humility’ ‘200 horses’ ‘his death’ ‘bottles’ ‘the army’ ‘all angels’ ‘Alexandria’ ‘wheat’ ‘water’ ‘ten years’

Named and Non-Named Entities Named Entity Non-Named Entity

“Alexandria” “A woman”

• •

Uniquely colored boxes Unique icon for each type

Named Entities also have • •

Yellow filled Boxes Wikipedia link in the corner

• •

Covers all entity types Broadest range of identifiers • Access to Wikidata API • Used by many other projects

“Linking” with other projects Projects that link to Wikipedia include: • • •

Pleiades Trismegistos

Using Wikipedia provides a common table of reference for many other projects studying the ancient world

Searching for entities We have ~250K words annotated for entities and automatic tools to add more. Explore entities in texts by using Coptic SCRIPTORIUM’s ANNIS. For more information and tips for using our tools visit: For a fuller explanation of entities see: Exhaustive Entity Recognition for Coptic: Challenges and Solutions.

Coptic SCRIPTORIUM is supported by the National Endowment for the Humanities Office of Digital Humanities and Division of Preservation and Access, Georgetown University, The University of Oklahoma, the University of the Pacific, and Canisius College.

Linked Pasts Symposium 2020, online, december 2020

PORTIC - a gazetteer for French ports [1749 - 1815] Dealing with imperfect data to publish an interoperable gazetteer

Christine PLUMEJEAUD-PERREAU 1, Mélissa MIMOUNI1, Alain BOUJU2 , Christian PFISTER3, Thierry SAUZEAU4, Silvia MARZAGALLI5

3. Université du Littoral-Côte d'Opale, ORCID 0000-0003-4802-942X 4. Criham (EA 4270), Université de Poitiers, France,, ORCID 0000-0001-8445-0807 5. CMMC (EA 1193), Université Côte d'Azur, France,, ORCID 0000-0003-0387-470X

1. UMR 7266 LIENSs, CNRS/Université La Rochelle, France., ORCID 0000-0001-9271-3355 2. L3I, Laboratoire d’Informatique de la Rochelle, Université La Rochelle, France,

PORTIC ( is an ANR program [2019 - 2023] which studies shipping activities and trade in France in the second half of the eighteenth century, insisting specifically on the visualisation of imperfect data. Through this research, we produce and publish on the semantic Web a gazetteer of French ports in activity between 1749 and 1815 and of foreign ports they were connected to, with a specification of their belonging to historical maritime and fiscal administrative districts, and to political entities. This work focuses in particular on how we dealt with imperfect data: the sources we used are lacunar and existing open data induce some uncertainties because they are not specifically devoted to maritime history. Result is published online : - Building a maritime gazetteer from French historical sources Geo_general – 1787 – 1789 : designed in 2010 as part of the Navigocorpus program, it serves as a major source to constitute our gazetteer. First based on the integration of 4 million of points provided by the National Geospatial-Intelligence Agency (NGA) of the United States, indicated by their longitude and latitude WGS84 and often endowed with several transliterations (e.g. London, London, Londra, etc.). 2063 Geo_general entries, i.e. 1250 distinct geolocated places identified by a local identifier (UHGS_ID) are retained and published in our gazetteer, of which 31% correspond to new manual entries. The work of historians in the Navigo corpus and then the PORTIC program has allowed to : - descend below the level of the municipality in order, for example, to distinguish the ports of Oneille and Port-Maurice, which are today two districts of the current municipality of Imperia, in Italy, whereas in 1787, Oneglia was not only a city apart, but it did not even belong to the Republic of Genoa, of which it was an enclave under the sovereignty of Piedmont; - be able to deal with localities all over the world; - take into account maritime points and landmarks; - take into account the old spellings, and their evolution over time, to be able to distinguish for example Saint-Nazaire, at the mouth of the Loire, from the homonymous locality in 1787 which is now called Sanary-sur-Mer (Var); - to associate the ports with political entities which no longer exist, such as the Papal States or the Ottoman Empire and their changes during the period [1749-1815]

Chardon – 1781 – 1785 : This exhaustive survey of French admiralties resorts has been used to attach each known port of Geogeneral to an admiralty, (and its French province subconsequently), to draw by hand limits of admiralties (see figure below), but also to discover new French ports that were not mentionned previously in Geo_general.

Zoom on French West atlantic coast

Imperfection in its geographic dimension The sources used do not allow you to know all the ports of France or the world at that time. Under its prism appear only the ports of admiralty and those which had a registrar to enter the taxes on ships, or ports intended for declared voyages. The superposition with alternative historical sources, (Cassini map, Chardon report), as well as current sources (GeoNames, World Port Index, SHOM ports base) makes it possible to complete the information and can provide an estimate of the missing information. However, the anachronism introduces new uncertainties (see map on right) :


About port existence : did the port of Île d'Aix exist in 1787?,

Geo_general Chardon or Cassini SHOM Port with a taxes office Limit of admiralties

About the precision of the location of the ports (that of La Pallice for instance), About multiple representations for certain ports such as Douarnenez

Belonging to territorial networks according to old geographical nomenclatures (admiralties, provinces, Fermes, states) is not easy to determine and still in progress for Fermes. This information is crucial for the analysis of flows grouped according to these divisions and the comparison with information from other corpora.

Quality controls: various certainty levels All locations were compared with entries in contempory gazetteers, like GeoNames, Wikidata, SHOM (French maritime database) and World Port Index database. The threshold for matching was set to 5 km, and we found a high matching rate for Geo_general with geonames, and much lower rates for WPI and SHOM, but the dispersion around an average distance of 1835 meters is high for all sources, which highlights the uneven quality of all sources. The distance to rivers or coasts has been computed in order to qualify the certainty of positions : a distance above 10 km is thus qualified as uncertain, as this is the case for 9% of the corpus. This is mostly due to the vagueness of the source that mentionned the destination of a ship as beeing a country or a region like Corsica for instance, and Corsica was located in the middle of the island. We have used the opportunity offered by the LinkedPlaces model ( to fill uncertainty levels for various elements and to date the period of belonging of entities inside the period [1749-1815]. Contemporary sources

Historical sources Geo_general


Cassini map – 1756-1790 was also screened to locate ports cited by Chardon and ports found on the map but never cited in other sources.

BD toponymes of SHOM

rate of matching

Mean distance (m)

Distance dispersion (%)




























Geo_general World Port Index

Number of ports
















Geographic distance in average (m)


Publish on the Web using the LinkedPast model This approach makes it possible to publish database online with an opensource API in October 2020, under the Open Data Base Licence. When the entities are ports, their geometry is a Point, when the entities are the admiralties, their geometry is a set of MultiLineString, and else, no geometry provided, because limits are not yet determined exactly for the provinces or the states. The geometries are saved in WGS84 geographic coordinates. Toponyms are mainly indicated in French, except for the names of states which are translated into English. The type of entity retained is therefore either a port, a notion which covers a very broad meaning, or an admiralty, or a province. The information is sourced, with a URI if possible (but there is not always a deferenceable identifier of the source). Part-of relations are all dated and sourced as well. A systematic link is given with the current GeoNames entity that best matches, by providing information such as identifying the URI of the entity in GeoNames ( for La Rochelle for example. Other links, to the TGN or wikidata may exist. The source code and documentation of the gazetteer is online : The gazetteer is online : You can get 12 provinces (aat: 300000774) description by their name : You can get 51 admiralties (aat: 300387179) description by their name : You can get 1250 ports (aat: 300120599) description by their UHGS id: A download all the database is available here :

What? 1791 Buddhist sculptures in the Dir Museum in Chakdara and in the Mission House of the Missione Archeologica Italiana in Pakistan in Saidu Sharif (Khyber Pakhtunkhwa, Pakistan): statues of the Buddha, bodhisattvas, tutelary deities, and donors, as well as narrative reliefs illustrating events of the Buddha’s life.

DiGA and Beyond Digitization and Linked Open Data for Gandhāran Art History Jessie Pons and Frederik Elwert Center for Religious Studies (CERES), Ruhr University Bochum


The artefacts stem from 13 ancient Buddhist archaeological sites located on the northern and southern banks of the river Swat: Andan Dheri, Bambolai, Barchoray, Chatpat, Damkot, Jabagai, Macho, Nasafa, Ramora, Sha Shamsi Khan, Shalizara, Talash and Saidu Sharif.


Making a new corpus available

Give more visibility to the multiplicity of Gandharan styles

Charting formal and iconographic features

Complement the map of production centers

Highlighting patterns of exchange

Enhance our perception of the development of the school

Reassessing interplay b/w images and texts

Clarify the image of Buddhism in the region

Buddhist Digital Resource Center Gandhara connection

DiGA DiGA project: Digitization of Gandhāran Art

Jātaka Stories


DiGA will ▪ be hosted as part of the heidICON database by CrossAsia/UB Heidelberg ▪ model the metadata following CIDOC-CRM and expose them via OAIPHM ▪ use GND as a reference vocabulary SPONSORED BY THE

Pelagios WG Linked Data Methodologies in Gandhāran Buddhist Art and Texts

The Pelagios WG ▪ identifies current standards in the field of cultural heritage that are relevant for Gandharan studies ▪ defines a set of vocabularies, creates new ones to fill possible gaps: Places (i.e. Buddhist archaeological sites), Motifs (i.e. decorative motifs, environment, objects, etc.), Narratives (i.e. episodes in the past and last existences of Buddha), Persons (i.e. historical, literary, archetypes) ▪ develops strategies to ensure interoperability and linking between databases. Report: doi:10.13154/rub.148.125

Setting up a comprehensive digital catalogue of the coins of Ancient Spain Pere Pau Ripollès Universitat de València

Manuel Gozalbes Museu de Prehistòria de València

Alejandro Peña, Juan Francisco Onielfa Render S.L.

MIB sections

Coin catalogue MIB ( is a website that will be launched in 2021. Its main goal is to arrange a catalogue with accurate descriptions for over 4,000 coin types minted in the Iberian Peninsula between 6th and 1st centuries BC. These coins were minted by Greeks, Punics, Iberians, Celtiberians and Romans. The project includes sections with detailed information for mints, hoards, finds, collections or auctions.

Software Cultural Heritage Management system

• Web application based on open standards (JSON, • • • • • • •

HTML, CSS, Javascript, PHP, PostgreSQL)

Open platform, open source

Browser based management system

Open to developers & contributors

Manage texts, images, pdfs and audiovisual materials

Multiuser, multiproject, multilanguage and multithesaurus

Cumulative sequential workflow

Time machine to recover deleted records and previous versions of field data

Linked data

Project facts

• The system includes 30,000,000 internal links establishing

40 Collaborators 8,000,000 Records edits 20 Data sections 1,000 Fields / components 255,000 Records 9,000 Thesauri terms 320,000 Images 30,000,000 Internal record links 100,000 Coins 4,000 Coin types 7,000 Publications 40,000 Bibliographic references

• •

relationships among the unique data of different sections. Managing linked values provides a powerful indexation and guarantees accurate searches into the website. The incorporation of linked resources is another common feature for every section. The website is designed for including unlimited links to the original providers of information; museums, libraries, journals publishers and numismatic projects like

The MIB website will offer stable URIs for the main sections allowing to link its contents in the right way. The URIs model is designed for preserving historical versions when changes in the contents occur.

Linked with libraries, museums, collections, journals, persons, places; Nomisma, Bnf, Oxford, British Museum, Zenon, Archeological

Distributed Text Services (DTS) A Community-driven Text Interoperability API

Specs: Demo: web + notebooks FAQ page

Linked Pasts 6, University of London and British Library, December 2-16, 2020

What is DTS? Goal: enable machine-actionable operations across a wide variety of TEI texts. As DTS strives for genericity, it does not dictate: how collections should be organized type of persistent identifiers to use - what ontologies to use for metadata - how the texts are structured.

Endpoints & Operations Collections endpoint: List available collections Support for pagination & virtual collections Navigate between nested collections Navigation endpoint: List available text passages Documents endpoint: Retrieve the text of a document


Multi-level/competing text hierarchies Any identifier scheme (*) Non-homogeneous text hierarchies


Hold metadata Serve as a catalog May be: nested readable chunked into citable textual elements

Passages: hold textual content

DTS provides a standard way for clients to interact with collections of TEI documents. A standard API allows users to access many text collections using the same client software. It also allows editors to publish text collections in a usable way that existing clients can use.

DTS supports:


References: hold identifiers & metadata for subsections of documents

Why DTS?

A Generic Text API

Classes of objects in DTS

DTS Community 46 Members

DTS-compliant resources APIs


ENC Alpheios Perseids EDH Beta maṣāḥǝft

TEI publisher client/server Perseids DTS API Nautilus + MyCapytain

Communication via Google Group list Technical exchanges: GH issues Minutes: meeting notes GH repo

Want to join? Subscribe to the Google Group

Crowdsourcing Chinese history: Distributed transcription, annotation, and datafication

Donald Sturgeon, Durham University

1. Transcription & Annotation • • • • • •

OCR + crowdsourced, versioned editing Real-time indexing Chunked XML representation Corrections during & after annotation Annotation via point-and-click UI Widely used open platform • • 5 billion characters of transcribed text • 30 million pages of scanned sources • 40,000 daily users

Transcription (OCR+crowdsourcing) ⼰卯,以鎮 為宜州觀察 使。賜⽟帶, 尋升慶遠軍 承宣使。…

18th century manuscript 2. Datafication • • • • • • •

Crowdsourced annotation

UI proposes claims based on annotations

Graph database; all annotations point to an entity Statement & qualifier data model (similar to Wikidata) Machine-readable dates Machine-readable textual references as evidence Point-and-click UI proposes, checks, and adds claims Claims augment knowledge base Augmented knowledge base used in subsequent annotation Annotation & extraction leverage entity data

Claims are added to entities with evidence

Linked Open Samian Ware - Transformation and modelling of vagueness and uncertainty

The online relational database `Samian Research` contains a quarter of a million finds of potters’ stamps on Terra Sigillata from all over Europe. These Gaulish potters’ stamps have been found on 4300 sites throughout the Roman Empire, allowing for dating sites and Roman trade research. In order to provide this resource as FAIR Linked Open Data, we created a semi-automatic transformation process as well as an CIDOC-CRM based ontology, starting from a PostgreSQL database resulting as a SPARQL endpoint in a RDF4J triplestore. In this process we also modelled vagueness and uncertainty for relations between actors and information carriers, actors and productioncentres as well as potforms and information carriers using the Academic Meta Tool. Florian Thiery M.Sc.

Dr. Allard Mees FSA

Dennis Gottwald B.A.

RGZM | RSE at Scientific IT ORCID | 0000-0002-3246-3531 E-Mail |

RGZM | Head of Scientific IT ORCID | 0000-0002-7634-5342 E-Mail |

JGU Mainz | Intern at RGZM Scientific IT ORCID | 0000-0002-8761-4659 E-Mail |

Entangled Ordinances – Linked Open Data for Legal History Dr. Andreas Wagner1 and Dr. C. Annemieke Romein2 1 Max-Planck-Institute

for European Legal History; 2 KNAW Huygens ING

In the 1990s the Max-Planck-Institute for European Legal History initiated a large study into police ordinances (Policeygesetzgebung) in 68 early modern states (light green on the map), leading to an inventory-dataset of metadata on over 200,000 texts. More recently, several independent initiatives have also inventorised ordinances in other areas (dark green). A hypothesis that could be answered, by using linked data, is which topics were copied from one area to another – or even if legislation was copied in its entirety. What you see on this poster, are our first attempts to come up with an ontology and a multilingual SKOS. Ideally, the ontology (and LOD) would integrate various levels of jurisdictions and norms (legislation).

For more info: • •

Early Modern Ordinances in the Dutch Republic and Dutch East India Company’s territories Dr. Annemieke Romein1; Vany Susanto, MA2 1. KNAW Huygens ING; 2. Leiden University

The Sphere of Influences of Dutch Ordinances

Placards/ ordinances were precursors of administrative legislation, covering a wide array of topics.

This genre was introduced to Asia by the European trading companies.

Initially, it was implemented to manage the Company’s men; but later also applied to the native inhabitants and local communities in Dutch Asia.

The texts applied to all inhabitants, without regard of birth and wealth. Gradually replacing the initial feudal hierarchy.

Ordinances in the Dutch Republic

Ordinances in the Dutch Asia

Ordinances within Europe/ the Republic addressed topics ranging from social order, public order, infrastructure, education/ health and economic affairs. A current inventory of the rules in Holland and Gelderland—linking these to the German categorization—could be linked to the Asian Ordinances to provide insights into the overseas exchange of the norms.

Initial study of the placards issued in Asia by the Dutch East India Company shows that the implementation was glocalized, meaning that the ordinances were issued to respond to the local environments, were negotiated with the local communities, and were issued in local languages.