University of San Diego University of San Diego
Digital USD Digital USD
Digital Initiatives Symposium
Apr 27th, 10:45 AM - 11:25 AM
Mapping Manuscript Migrations: Building and Using a Linked Mapping Manuscript Migrations: Building and Using a Linked
Open Data Environment for Medieval and Renaissance Open Data Environment for Medieval and Renaissance
Manuscript Studies Manuscript Studies
Lynn Ransom University of Pennsylvania, [email protected]
Toby Burrows University of Oxford, [email protected]
Follow this and additional works at: https://digital.sandiego.edu/symposium
Part of the Cataloging and Metadata Commons, Digital Humanities Commons, European History
Commons, and the Medieval Studies Commons
Ransom, Lynn and Burrows, Toby, "Mapping Manuscript Migrations: Building and Using a Linked Open Data Environment for Medieval and Renaissance Manuscript Studies" (2021). Digital Initiatives Symposium. 13. https://digital.sandiego.edu/symposium/2021/2021/13
This Event is brought to you for free and open access by Digital USD. It has been accepted for inclusion in Digital Initiatives Symposium by an authorized administrator of Digital USD. For more information, please contact [email protected].
Mapping Manuscript Migrations: Building and Using a Linked Open Data Mapping Manuscript Migrations: Building and Using a Linked Open Data Environment for Medieval and Renaissance Manuscript Studies Environment for Medieval and Renaissance Manuscript Studies
Presenter 1 Title Presenter 1 Title Curator, SIMS Programs
Presenter 2 Title Presenter 2 Title Senior Researcher
Session Type Session Type Event
Abstract Abstract “Mapping Manuscript Migrations” is a digital humanities project that brings together three distinct data sets about the histories of more than 215,000 medieval and Renaissance manuscripts for browsing, searching, and visualization. Four leading institutions from Great Britain, France, Finland, and the United States collaborated on this project, pooling their expertise in Semantic Web technologies and medieval manuscript curation and research, as well as contributing their own data from the three contrasting datasets. The Schoenberg Database of Manuscripts at the University of Pennsylvania, the Medieval Manuscripts Catalogue at the University of Oxford, and the Bibale database from the Institut de recherche et d’histoire des textes--are brought together in a Linked Open Data environment, constructed by the team members from the e-Research Centre at the University of Oxford and the Semantic Computing Group at Aalto University in Finland, to aggregate, enhance, and present the data, with a data model based on the CIDOC-CRM and FRBROO ontologies.
While also considering the challenges and successes of this international collaboration, Dr Lynn Ransom (Schoenberg Institute for Manuscript Studies, University of Pennsylvania) will show how the project builds on the data and functionality of the source datasets and enables new approaches to research in manuscript history and provenance.
Location Location
Keywords Keywords Linked open data, manuscript studies, provenance studies, databases, data modeling, semantic web, interfaces, rdf query, SPARQL
Creative Commons License Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License
This event is available at Digital USD: https://digital.sandiego.edu/symposium/2021/2021/13
Dr Toby Burrows Dr Lynn RansomOxford e-Research Centre Schoenberg Institute for Manuscript StudiesUniversity of Oxford University of Pennsylvania
@MSMigrations @sims_mss
Mapping Manuscript Migrations: Building and
Using a Linked Open Data Environment for
Medieval and Renaissance Manuscript Studies
Mapping Manuscript Migrations Project Team
• Oxford: Toby Burrows, David Lewis, Andrew Morrison, Kevin Page, Pip Willcox, Thanasis Velios, Graham Klyne
• Aalto: Eero Hyvönen, Esko Ikkala, Mikko Koho, Jouni Touminen
• IRHT (Paris): Hanno Wijsman, Nicole Bergk Pinto, Antoine Brix, Mahaut Cazals, Alexandre Gaudin, Synnøve Myking, Pierre-Louis Pinault, Guillaume Porte
• Pennsylvania: Lynn Ransom, Doug Emery, Mitch Fraas, Benny Heller, Emma Thomson
Boccaccio, GiovanniDe Claris MulieribusFrance, probably Besançon ca. 1464-1470
Sold at Christie’s New York, 2001 US$47,000
Travels of the Boccaccio MS ca. 1464/70 to 2001
6
7
Data Sources
Medieval Manuscripts in Oxford Libraries
Text Encoding Initiative (TEI) XML documents
10,000 catalogue records; about 1,000 with detailed provenance information
Schoenberg Database of Manuscripts
Relational database - bespoke data model
240,000 records containing provenance-related observations of manuscript histories
BibaleRelational database - bespoke data model
More than 16,000 manuscript records with detailed provenance information
8
Linked Data 1
Slide: Dr Kevin Page (OeRC)
Linked Data 2
Slide: Dr Kevin Page (OeRC)
Linked Data 3
Slide: Dr Kevin Page (OeRC)
Linked Open Data Stack
MMM Project Products• Unified Data Model for manuscript history and provenance (derived from
CIDOC-CRM and FRBR ontologies)
• Conversion and transformation tools and pipelines
• LOD vocabularies with identifiers for 222,000 manuscripts, 435,000 works, 56,000 actors, 5,000 places
• Public SPARQL endpoint to the MMM triple store (20 million RDF triples)
• Public portal: browsing, searching, visualizations (Sampo-UI software)
• Data export and downloads (from the public portal and Zenodo data repository)
• GitHub site for tools, data, and documentation
• Publications and presentations (50+)
MMM Project Outputs
MMM Data Model
MMM Data Model
Reconciliation: Vocabularies
Data Transformation Pipeline
Data Transformation Pipeline
Diagram: Dr Mikko Koho (Aalto University)
Access to the Mapping Manuscript Migrations data
222,000 Manuscripts 435,000 Works937,000 Events 5,000 Places 56,000 People and Organizations (Actors)
• Browse and search – via the MMM Portal (Sampo-UI software)
• Visualize the data on maps: places of origin; migrations over time; last known location – via the MMM Portal
• Inspect the underlying data – via Linked Data Finland
• Explore the data with SPARQL queries – via SPARQL endpoint + Yasgui
• Export the data into your own software environment – via Zenodo (all data) or MMM Portal + Yasgui (subsets of data)
19
20
21
22
23
24
25
26
27
Using MMM data for research
• Which French collectors acquired manuscripts in the century after the end of the Wars of Religion (1598)? Where are their manuscripts now?
• Which manuscripts containing texts by Ramon Llul were sold in the 19th century?
• Which manuscripts formerly owned by Sir Thomas Phillipps (1792-1872) are now in North American libraries?
• How many manuscripts were produced in London in the 15th century?
• Did Sir Thomas Phillipps own a thirteenth-century bible with historiated initials?
Original MMM Research Question
Bibale results – public interface
Bibale interface – comments
Oxford results – public interface
Oxford interface – comments
SDBM results – public interface
SDBM interface – comments
What French collectors purchased manuscripts in the century after the end of the Wars of Religion (i.e., after 1598)? Where are their manuscripts now?
Currently impossible. One can currently not run a query on transactions of a specific period (e.g. after a specififc year).
One general problem (in Bibale, in the UI and elsewhere) remains that it is difficult to define a "French collector", since it can involve birthplace, but also any other place of activity. We have discussed this at several occasions.
You can filter the list of "People" by role (e.g., "Owner, signer, or donor”). But you cannot further filter this list by country - whether this is place of birth, place of death, or place of residence.
The interface does not display any information about individual people (dates or places of birth and death, etc.).
This question can't easily be answered. You can search for SDBM Names that are linked to France and have life dates after 1598, but then you'd have to run separate queries to find people linked to places nested within France (there's currently no way to search for France and all of its children within the same query). Once you found all of these names, you could then view the entries linked to them and determine last known locations, but this would be very time consuming.
SDBM could be improved by allowing for nested place searching (the data is in there, it just isn't possible in the UI).
32
Really Digging Into the Data:
Querying the MMM SPARQL Endpoint
33
34
SPARQL Endpoint
https://api.triplydb.com/s/mvOuzPWvk
New Research Questions?
35
New Research Questions?
Which collectors bought manuscripts from Wilfrid Voynich?
36
CULTIVATE MSS Project: Cultural Values and the International Trade in Medieval
European Manuscripts, c. 1900-1945
https://www.ies.sas.ac.uk/research-projects-archives/cultivate-mss-project
New Research Questions?
Which collectors bought manuscripts from Wilfrid Voynich?
Where were the collectors located?
37
CULTIVATE MSS Project: Cultural Values and the International Trade in Medieval
European Manuscripts, c. 1900-1945
https://www.ies.sas.ac.uk/research-projects-archives/cultivate-mss-project
New Research Questions?
Which collectors bought manuscripts from Wilfrid Voynich?
Where were the collectors located?
What do we know about the kind of manuscripts he sold?
38
CULTIVATE MSS Project: Cultural Values and the International Trade in Medieval
European Manuscripts, c. 1900-1945
https://www.ies.sas.ac.uk/research-projects-archives/cultivate-mss-project
New Research Questions
Which collectors bought manuscripts from Wilfrid Voynich?
Where were the collectors located?
What do we know about the kind of manuscripts he sold?
https://api.triplydb.com/s/E6ed-lb0a 39
(etc …)
● 80 transfer events of manuscripts sold by Voynich between 1900-1934● Sold to mostly American and European individuals and institutions● Manuscripts produced in western Europe and England, ranging in date
from the 10th to 15th century
https://api.triplydb.com/s/E6ed-lb0a
etc...
Results
https://api.triplydb.com/s/wRZLM-hR4
New Research Question
What is the average height to width ratio for liturgical manuscripts, specifically missals, breviaries and the quasi-liturgical books of hours?
https://api.triplydb.com/s/wRZLM-hR4
Results
● Height to width ratios for a majority of liturgical manuscripts (missals, breviaries, books of hours) average around 1:5
● Outliers suggest non-traditional formats or bad source data
New Research Question
What can we learn about the social backgrounds of 19th-century English manuscript collectors?
https://api.triplydb.com/s/D1ITlSyTW
New Research Question
What can we learn about the social backgrounds of 19th-century English manuscript collectors?
https://api.triplydb.com/s/D1ITlSyTW
Wikidata search
Results
https://api.triplydb.com/s/D1ITlSyTW
etc ...
● 89 distinct names of people with death dates in the 19th century, who have a total of 81 different occupations between them.
● Politicians (21) and writers (17) are the most common, though there are also an astrologer, a brewer, and at least three slave holders.
● Each person has an average of two occupations, though some have many more than this: 18 in the case of William Morris!
New Research Question
How many women were involved in collecting between 1900-1950 and what were their social backgrounds?
https://api.triplydb.com/s/vnxpNsDQJ
Results
https://api.triplydb.com/s/vnxpNsDQJ
● 8 distinct women● diverse range of careers, from art collector to statistician, lawyer, economist, etc.● Low number of female collectors
Results
https://api.triplydb.com/s/vnxpNsDQJ
● 8 distinct women● diverse range of careers, from art collector to statistician, lawyer, economist, etc.● Low number of female collectors
Why?
Results
https://api.triplydb.com/s/vnxpNsDQJ
● 8 distinct women● diverse range of careers, from art collector to statistician, lawyer, economist, etc.● Low number of female collectors
Why?● Low number of female collectors that have been given VIAF numbers in both MMM and
Wikidata
Manuscript provenance data – future directions
• Improving the structure and content of provenance data in library catalogues and databases, e.g., MARC and TEI
– Last-known location versus current location
– Mapping ownership chains for a specific manuscript
– Specifying distinct types of events
• Working towards Linked Open Data for medieval and Renaissance studies
– Deciding which external vocabularies are most effective for reconciliation
– Lack of specialist medieval and Renaissance vocabularies in Linked Open Data form
– Implementing unique global identifiers for manuscripts (ISMI)
• Making data available for reuse– Formats, licencing, platforms
Mapping Manuscript Migrations
MMM Portal: https://mappingmanuscriptmigrations.org/
More information: http://blog.mappingmanuscriptmigrations.org/
https://github.com/mapping-manuscript-migrations
Twitter: @MSMigrations
Contacts:o Dr Toby Burrows [email protected]
o Prof. Eero Hyvönen [email protected]
o Dr Lynn Ransom [email protected]
o Dr Hanno Wijsman [email protected]