Biblissima - Following Medieval Manuscripts and Incunabula through their Existence via a Semantic Web Application

Biblissima—Bibliotheca bibliothecarum novissima—is a Digital Humanities project that brings together libraries and research institutions, whose goal is to provide a single access point to over 40 databases regarding medieval and Renaissance manuscripts and early printed books by the end of 2016.
The project began late in 2012 with nine founding partners: the Department of Manuscripts at the BnF (Bibliothèque nationale de France), the Campus Condorcet, the CESR (Centre d'Etudes Supérieures de la Renaissance), the CIHAM (Histoire, Archéologie, Littératures des mondes chrétiens et musulmans médiévaux), the Centre Jean-Mabillon at the ENC (École nationale des chartes), the CRAHAM (Centre de Recherches Archéologiques et Historiques Anciennes et Médiévales. Centre Michel de Boüard), the IRHT (Institut de recherche et d’histoire des textes), the Digital Document Centre at the MRSH de Caen (Maison de la Recherche en Sciences Humaines), and the SAPRAT-EPHE (Savoirs et pratiques du Moyen Âge au XIXe siècle, Ecole pratique des hautes études).
Funded by the French National Research Agency (ANR: Agence Nationale de la Recherche), Biblissima is concerned with the history of collections and the transmission of texts. In this respect, the project stands somewhat in line with the works of Antoon Sanders, who published his Bibliotheca Belgica Manuscripta in 1641-44, and Bernard de Montfaucon and his Bibliotheca bibliothecarum manuscriptorum nova, published in 1739, both of which are major inventories compiling lists of manuscripts held in many different libraries in Europe.
In order to create an Online library of historical collections of France for the 21st century, our chosen solution was to develop a semantic web application. We released a prototype (http://demos.biblissima-condorcet.fr/prototype) in summer 2015 that provides unified access to a subset of two major iconographic databases: Initiale (http://initiale.irht.cnrs.fr/accueil/index.php) and Mandragore (http://mandragore.bnf.fr/html/accueil.html). An export of the metadata pertaining to the illuminations that depict geographical locations was used as a starting point for the development of our application. People (author or illuminator), institutions (libraries), work titles and places were aligned with the BnF authority records and, whenever possible, also linked to external vocabularies (GeoNames, VIAF, Pleiades, Getty thesauri).
The next step involved generating dynamic web pages that describe a person, a work, an expression, a manuscript, a part of a manuscript or an illumination, and which include links to the corresponding descriptions in the source databases. This original data is enriched with longitude and latitude coordinates for geographical names that were acquired by aligning them with the equivalent GeoNames concepts. It is now possible, for example, to show on two different maps the places depicted in illuminations with the images of the corresponding folios, and the illuminations from manuscripts held in a particular institution.
In addition, a viewer embedded in the page (we currently favour Mirador: https://iiif.github.io/mirador) shows the digitised manuscript or folio when available. The information needed to display these images is structured according to the Shared Canvas data model (http://iiif.io/model/shared-canvas), which is the foundation of IIIF (International Image Interoperability Framework: http://iiif.io). This information is passed to the viewer in form of JSON-LD manifests, which are generated from metadata and image files supplied by the partner institutions. The viewer itself is client-based and features a deep-zoom capability for the loaded images, the possibility to display associated metadata, and also the superposition of different layers, such as an image and the corresponding textual transcription.
The data model behind the application is based on CIDOC-CRM (http://www.cidoc-crm.org) and FRBRoo (http://www.cidoc-crm.org/frbr_inro.html) and our data is already available in RDF format. Beyond taking into account the different levels of work, manifestation and item (using the classes F4 Manifestation Singleton for manuscripts as well as F3 Manifestation Product Type and F5 Item for early printed books), we intend to group manuscripts and early printed books as productions of certain expressions (class F2 Expression). Illuminations are understood as features (class E26 Physical Feature) placed on a folio or page, and provenance marks will be modelled in the same way. Based on the lessons learned through developing our own ontology, we will be able to give feedback about the data.bnf.fr, FRBRoo and SharedCanvas ontologies, as well as other theoretical foundations.
We are currently working on extending this prototype, which was developed using the semantic web application framework CubicWeb (https://www.cubicweb.org/). In addition to manuscripts, we will be including early printed books, with a strong emphasis on book provenance, by integrating data from the following four sources: Esprit des Livres (database on auction and other sales catalogues, ENC), Bibale (database on historical book collections, IRHT), Europeana Regia (database on three important historical collections of the Middle Ages and Renaissance), as well as CR2I (Catalogues Régionaux des Incunables Informatisés, CESR) and CRIICO (CR2I Centre-Ouest, CESR). Resolved challenges include the creation of ARKs (Archival Resource Keys) for each manuscript (BnF and IRHT) and codicological unit (BnF). At a later stage, digital editions of inventories of manuscripts in TEI XML will be integrated as well.
This short paper will present the latest version of the semantic web application for Biblissima’s data cluster.