=Paper=
{{Paper
|id=Vol-3160/paper10
|storemode=property
|title=RESTORE: Opening Data in Digital Humanities and Cultural Heritage
|pdfUrl=https://ceur-ws.org/Vol-3160/paper10.pdf
|volume=Vol-3160
|authors=Francesco Coradeschi,Emiliano Degl'Innocenti,Carmen Di Meo,Maurizio Sanesi,Alessia Spadi,Federica Spinelli
|dblpUrl=https://dblp.org/rec/conf/ircdl/CoradeschiDMSSS22
}}
==RESTORE: Opening Data in Digital Humanities and Cultural Heritage==
https://ceur-ws.org/Vol-3160/paper10.pdf
RESTORE: Opening Data in Digital Humanities and Cultural
Heritage
Francesco Coradeschi, Emiliano Degl’Innocenti, Carmen Di Meo, Maurizio Sanesi,
Alessia Spadi and Federica Spinelli
Consiglio Nazionale delle Ricerche, Istituto Opera del Vocabolario Italiano, Firenze, Italy
Abstract
The RESTORE project (smaRt accESs TO digital heRitage and mEmory) started in June 2020
with a duration of 2 years. The project consortium, coordinated by the Istituto Opera del
Vocabolario Italiano of the Italian CNR (National Research Council of Italy), includes national
Cultural Heritage institutes, such as the State Archives and the Museum of Palazzo Pretorio in
Prato and the Archival and Bibliographic Superintendency of Tuscany, and the SPACE SpA
software company. The project - co-financed by the Regione Toscana - has its main purpose
in the recovery, integration and accessibility of data and digital objects collected by partner, in
order to build a knowledge base made of information regarding the history of the city and of
its civic institutions, the development of its economic and entrepreneurial system, the role of
women in the development of a welfare state and network. Starting a local history approach, it
is nonetheless possible to broaden the focus from the local dimension to reconstruct a
significant part of the history of European and Mediterranean cities of the 14th century,
including commercial and economical aspects. This paper presents a focused overview on the
mapping and modeling of archival and museum digital resources encoded with different
standards, such as XML-EAD, XML-EAC, XML-TEI and ICCD-OA, and covered in the
CIDOC-Conceptual Reference Model, the ontology that was chosen as “common language”
for semantic data integration. The long term sustainability for the project’s results will be
fostered by the collaboration with key players in the EU RIs environment, such as DARIAH-
ERIC (ESFRI Landmark for the Humanities and Social Sciences) and E-RIHS (ESFRI project
for the Heritage Science), as well as other actors within the EOSC Framework.
Keywords 1
FAIR, collections, archives, GLAM, ontologies, CIDOC-CRM, conceptual modeling,
mapping, data lifecycle, Linked Open Data, semantic Web, metadata integration, cultural
heritage, semantic data, knowledge management, social sciences and humanities.
1. Introduction23
RESTORE (smaRt accESs TO digital heRitage and mEmory1), is a project coordinated by the CNR-
OVI2 (Istituto Opera del Vocabolario Italiano, Consiglio Nazionale delle Ricerche) based in Florence.
The goal of the RESTORE project consists in the creation of a digital environment for data integration,
with a user-friendly graphic interface focused on datasets coming from the domains of Digital
IRCDL 2022: 18th Italian Research Conference on Digital Libraries, February 24–25, 2022, Padova, Italy
fr.coradeschi@gmail.com (F. Coradeschi); emiliano.deglinnocenti@cnr.it (E. Degl’Innocenti); dimeo@ovi.cnr.it (C. Di Meo);
maurizio.sanesi.eng@gmail.com (M. Sanesi); spadi@ovi.cnr.it (A. Spadi); spinelli@ovi.cnr.it (F. Spinelli)
0000-0003-0808-2736 (F. Coradeschi); 0000-0002-3839-9024 (E. Degl’Innocenti); 0000-0001-9953-7753 (C. Di Meo);
0000-0003-3917-5935 (M. Sanesi); 0000-0002-9670-3426 (A. Spadi); 0000-0002-2195-3930 (F. Spinelli)
©️ 2022 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
1
The project website is available at: http://restore.ovi.cnr.it
2
OVI-CNR: http://ovi.cnr.it
Humanities and Heritage Science. The data integration platform will also allow users to extract relevant
entities and other valuable information from datasets originally encoded by the partners in a variety of
standards in a new integrated digital environment. The platform relies on a modular architecture -
integrating custom components realized by the RESTORE dev team with existing solutions already
available - to support the conversion of the digital resources provided by the project partners, to the
CIDOC-CRM ontology. Among the supported formats: TEI3 for texts; EDM4 and MAG5, MODS,
METS6 for content produced by libraries; EAD7 and EAC8 for archives; ICCD-OA9 for museum
collections; formats in use in the heritage science like EDF10, HDF511. The final version of the
RESTORE platform will also include components for data ingestion and storage (i.e. CKAN12) data
management, access and integration (based on the Virtuoso13 triplestore). 4567891011121314
Starting from a keen analysis of the original datasets, taking into account the specific production
context (e.g. libraries, archives, museums, etc.) and the integration requirements expressed by the
partners, the team designed and implemented a set of custom solutions, including tools for the
normalization, syntactic alignment and semantic modeling of data. using the CIDOC-CRM14 as the
reference ontology, allowing the representation of both tangible and intangible aspects of cultural
heritage artifacts. 15
In this contribution, the integration methods so far implemented are presented, with a focus on the
sources provided by the Archivio di Stato di Prato15, the Museo di Palazzo Pretorio16 and the Istituto
Opera del Vocabolario Italiano (cfr. note 3). The target audiences for the project are represented by
experts and researchers specialized in the scientific domains of reference; active citizen scientists
involved in cultural institutions’ activities; average people who want to get access to the consultation
of data. 1617
The project follows a FAIR17 [1] approach to data integration, and is developed in line with the
guidelines for data management provided by EU projects and other international actors promoting Open
Science for the SSH domains, such as the H2020 SSHOC18- Social Sciences and Humanities
Open Cloud thematic cluster, the DARIAH-EU19 research infrastructure, ESFRI20 Landmark for Digital
Humanities; and the E-RIHS21 research infrastructure, ESFRI Project for Heritage Science. 1819202122
2. Original resources and their data structure
In this paragraph, a synthetic overview of material and sources that the team can ingest and process is
given. It comprises, as samples of the workflow, archival fonds, museum collections, textual material
from libraries and research centers.
Archival fonds and resources:
3
Text Encoding Initiative - TEI: https://tei-c.org/
4
Europeana Data Model - EDM: https://pro.europeana.eu/page/edm-documentation
5
Administrative and Management Metadata - MAG: https://www.iccu.sbn.it/export/sites/iccu/documenti/manuale.html
6
Metadata Object Description Schema - MODS e METS: https://www.loc.gov/standards/mods/presentations/mets-mods-morgan-ala07
7
Encoded Archival Description - EAD: https://www.loc.gov/ead/
8
Encoded Archival Context - EAC: https://eac.staatsbibliothek-berlin.de
9
Istituto Centrale per il Catalogo e la Documentazione - ICCD: http://www.iccd.beniculturali.it/
10
European Data Format - EDF: https://www.edfplus.info/
11
Hierarchical Data Format - HDF: https://www.hdfgroup.org/solutions/hdf5
12
Comprehensive Knowledge Archive Network - CKAN: is an open source tool that allows the creation of websites based on open data and
the publication of datasets, making them accessible to multiple users. To learn more, see the dedicated website: https://ckan.org/
13
Virtuoso Open Link: https://virtuoso.openlinksw.com
14
CIDOC - Conceptual Reference Model: http://www.cidoc-crm.org/
15
Archivio di Stato di Prato: http://archiviodistato.prato.it
16
Museo di Palazzo Pretorio: http://www.palazzopretorio.prato.it/
17
Findability, Accessibility, Interoperability and Reuse - FAIR: https://www.go-fair.org/fair-principles/
18
Social Science and Humanities Open Cloud - SSHOC: https://www.sshopencloud.eu/
19
Digital Research Infrastructure for the Arts and Humanities - DARIAH-EU: https://www.dariah.eu/
20
ESFRI: https://www.esfri.eu/esfri-roadmap-2021
21
European Research Infrastructure for Heritage Science - E-RIHS: http://www.e-rihs.eu/
- the “Datini22” collection is the largest mercantile archive for the Middle Ages available,
consisting of 150,000 letters and about 600 registers from which it is possible to extract information on:
i) people; ii) costs and types of goods involved; iii) places mentioned, etc.; the “Ospedale Misericordia
e Dolce23” collection, with its 7000 archival units, presents all the articulations of the functions of an
assistance institution: from support to the wayfarer, to the care of the poor and the sick, up to the
reception of the children abandoned and reared, thanks to the hospital itself, by the entire community
of Prato; Museum collections: 2324
- a selection of metadata regarding the works of art with a connection to the “Ospedale
Misericordia e Dolce” (see above). These are preserved as part of the artworks collection held by the
Museo di Palazzo Pretorio di Prato, and cataloged according to the ICCD standard (i.e. “OA” - “work
of art” entries); Textual corpora and text based resources:
- lemmatized correspondence made available by the CNR-OVI Institute. The dataset contains
3012 letters from the Datini Archive and selectively lemmatized. The lemmatized elements are
distributed within 22 categories (hyperlemmas) which include, in addition to personal and place names,
terms and verbs pertaining to the religious field, agriculture, parts of the body, names of the week, the
generic terms month and year, including: clothing and furnishings, food, animals, arts and crafts,
calendar, economics, law and politics, construction and architecture, medicine, coins, navigation,
kinship, leather goods and textiles, etc.2425
3. Workflow
The aim of the project is to achieve full data integration for the project’s partners datasets produced
following various procedures, normative schemas, standards and cataloging systems - and with different
disciplinary concerns. It is, therefore, necessary to set up a workflow that not only allows the
semantization of data within a single reference domain, but also proposes a shared data processing
model, valid for multiple metadata schemes and for multiple standards. On the technical side, it was
necessary to clean the data before processing, to remove elements resulting from incorrect data
processing (i.e. use of different management tools, standards and styles) undermining its understanding.
The technical phase is accompanied by a periodic focus with domain experts, to carefully analyze
the materials to be treated and the effectiveness of the solutions provided by the team. Following the
feedback provided, the team aims to implement a navigation environment that holds in itself the
semantic meanings discussed with domain experts. This is possible by structuring the data as a semantic
dataset and by defining the relationships that exist between entities. 2627
As a first step taken in data management, the original resources were uploaded to a data store, for
storage and access: the team has activated an instance of the CKAN tool (cfr. note 14) in which the
original XML files, i.e. the sources provided by the partner institutions, are stored and described. In
fact, CKAN allows for the description of datasets by means of an association of the original metadata
files to its own metadata system (so, it essentially results as a description of the data collection). The
project workflow involves the transformation of the data provided, regardless of the original format,
into semantic triples25. A “triple” is a statement on the semantic data, expressed using a set of three
elements: Subject, Object and Predicate; the triples are the smallest building blocks for the RDF 26
22
Collection of administrative documents and correspondence of the merchant Francesco di Marco Datini
(1335-1410) which testifies, through his vast activity in the industrial, commercial and banking fields, a cross-section of the economy and
social life of the entire Mediterranean basin.
23
A charitable organization that has cared for wayfarers, the poor, and abandoned children since the 13th
century. The digital resources related to this fund can be consulted on the website of the Archives: http://www.archiviodistato.prato.it/accedi-
e-con-sulta/aspoSt005/tree
24
The lemmatised corpus of the Datini correspondence, produced by CNR-OVI between 2003 and 2005,has been developed with the software
Gestione degli Archivi Testuali del Tesoro delle Origini - GATTO (http://www.ovi.cnr.it/Il-Software.html), the same programme that manages
the corpus Tesoro della Lingua Italia delle Origini - TLIO (http://tlio.ovi.cnr.it/), in a specially dedicated version that can be queried online
via the GattoWeb software (http://aspweb.ovi.cnr.it/(S(qmmiy5m0sybb4lao4qqmexyo))/CatForm01.aspx)
25
See: https://www.w3.org/TR/rdf11-primer/#section-triple
26
Resource Description Framework - RDF: https://www.w3.org/RDF/
format. A triple, therefore, is the result of an arrangement of the data itself in the form of subject-
predicate-object expressions, where the different components of the expression can be associated with
a single and unique URI27, or sequences of characters specifically identifying the resources that have
been mapped and “transferred” in a semantic context. The conversion of the original data into triples
was carried out through the use of custom parsers28. 2829
The first parser of the workflow takes in input the original data sources, and outputs them in the form
of tables results encoded in a CSV29 format, following a 1:1 ratio (the corresponding CSV file is
produced for each original file). So, the program processes the original data by ordering them in a
predetermined way, corresponding to the display of the same aggregated data in tabular form. The next
step in data processing involves transforming the information contained in the CSV tables into triples,
in the TTL30 format. This phase is also achieved through the use of a parser implemented ad hoc in
Python. Data triplication of the data, or the conversion of the same from the table format (CSV) into
triples (TTL), implies a further analysis of the data, for it is based on the nature of the information to be
conveyed. It is, therefore, necessary to make the following operations: i) choosing the most relevant
information (with the help from experts in the relevant scientific domain); ii) specifying the meaning of
the information to be represented (depending on the context of use, etc.); iii) providing for the actual
modeling in CIDOC-CRM, choosing the appropriate entities and properties. 3031
In the context of the activities described, the mapping indicates a process of aligning information of
the same kind, though contained in different datasets, and expressed through different semantic schemes
and/or structures - through the combination of elements belonging to one dataset with those of another
dataset if both express the same concept (eg: signature, author, place, etc.).
The data mapping phase facilitates the migration of data to be converted into logical constructs
articulated on the basis of the CIDOC-CRM model. The ontology provides the developers with a model
of management and semantic representation of knowledge bases allowing for defining a set of concepts
and describing fundamental relationships between them. The finality of the process is to formalize the
knowledge (i.e.: translate it into a formal language31), so as to minimize its ambiguity32 (typical of
natural language) and make it “computable”, machine readable, through a module called reasoner, so
the calculator can make inferences about it. Through the modeling, the data becomes structured
according to the rules of the reference ontology, which will form the basis for the transformation of the
datasets into triples. 3233
To proceed with the mapping and modeling it was necessary to conduct an in-depth study of the
EAD and EAC standards archival data encoded with such a standards [2][3]; an in-depth study of the
ICCD regulations (OA forms, regulation 3.00, 2018) for the entries regarding the collections of works
of art [4]; an in-depth study of the TEI standard for the representation of texts in digital format [5]; the
study and selection of the reference ontology which could better be used as a semantic basis to build
the triples. Through the use of the common ontology, the data were integrated and aligned to the
CIDOC-CRM entities and properties. Of particular interest are the entities that in the mapping phase
were identified as possible points of alignment between the different datasets, present in all the
resources: persons, places, objects. The effective integration of the data, is based on the reference to a
unique identifier (ie: a unique and possibly persistent URI) for the entities common to multiple datasets
(ie: the toponym Prato, or the anthroponym Francesco di Marco Datini), which constitutes the point of
contact between competing or ambiguous descriptions, and/or different information relating to the same
objects (i.e.: people, places, etc.) coming from different information contexts (archives, museums,
libraries, etc.).
4. Mapping and Modeling
27
Universal Resource Identifier - URI: https://www.w3.org/wiki/URI
28
For the generic definition of parser, see the Treccani online: https://www.treccani.it/enciclopedia/parser/. In this specific case, parsers are
programs in the Python programming language for transforming data from one format to another.
29
Comma Separated Values - CSV: https://datatracker.ietf.org/doc/html/rfc4180
30
Turtle - TTL: https://www.w3.org/TR/turtle/
31
cfr: Gottlob Frege, Begriffsschrift, eine der arithmetischen nachgebildete Formelsprache des reinen
Denkens, Halle a. S.: Louis Nebert, 1879: https://gallica.bnf.fr/ark:/12148/bpt6k65658c
32
cfr: Bertrand Russell, On Denoting, Mind 14, no. 56 (1905): 479–93: http://www.jstor.org/stable/2248381
After a deep analysis of the standard used by the different institutions that provided the database of
the project we started the modeling activity from the elements identified as conjunction points between
the collections at disposal, following the chosen ontology, CIDOC-CRM. In addition to the ontology (a
scheme) it was also necessary to select a series of reference vocabularies, for each domain of interest
(authors, artists, places, etc.). Through the references to vocabularies it was possible to identify and
index toponyms and anthroponyms, two of the entities identified as points of contact between the
various datasets, and general concepts both materials and immaterials. The common entities that occur
in the different datasets will therefore have to share the link to the reference vocabulary. In this way,
the integration of data on entities that refer to the same anthroponym or toponym is achieved. In fact, if
entities belonging to different datasets refer to the same URI, they will be automatically linked also
through the same vocabularies.
4.1. Defining objects and identifiers
The standards studied and described in this paper concern the description of handmade things. From
the CIDOC-CRM perspective, a handmade thing is a physical object purposely created by human
activity. For this reason, the E22 Man-Made Object CIDOC-CRM class has been used for representing
the object, which the information refers to. Actually, the alignment to the CIDOC-CRM ontology made
it necessary to distinguish between two levels of understanding in regards to the description of the
objects treated. The artifacts (man-made objects, such as texts, works of art, written records of any kind)
were, in fact, considered both as a product of intellectual work - as an “idea” (the stages of its
commission and design by the author/artist) - and as physical, tangible manifestation of the same type
of work (as elements forming part of a physically identifiable collection and, therefore, placed,
musealised or archived, protected and even kept by organizations and “personas'' throughout history).
The first level designates the immaterial (intangible) aspect of the work from the moment of its
creation (even if only conceptual), while the second level describes its production and transposition
onto a physical (tangible) support. The distinction of the two aspects is indeed important, because those
different points of view equally contribute to the realization of the object as a whole; for the purposes
of data mapping, this fact imposes a reflection on the information to be circumscribed and described,
requiring a correct, mutual and concrete understanding. The two levels just distinguished are therefore
expressed through the CIDOC-CRM ontology, as follows:
● E73 Information Object: information object (idea)
● E22 Man-Made Object: physical object (reification)
The physical object refers to all the physical and tangible characteristics of the work, while the
information object refers to all the conceptual information.
As a physical object with a physical mass the resource has a physical location, expressed through
unique identifiers described by catalogs and archives and assigned to resources. These identifiers are
identified respectively in the signature for documentary materials preserved in archives and libraries
and in the Unique Property Code (NCT) for artworks.
Traditionally, the shelfmark of archival documents is composed of the name of the fonds and the
identifier of the archival unit, possibly divided into several sections (envelope, insert or file, code of the
documentary unit). On the other hand, artworks are cataloged according to the ICCD (OA schema 33)
regulations with a unique identification code (NCT): this is a number given by the sequence of values
assigned to the subfields Region Code (NCT+R) and General Catalog Number (NCT+N), which is
assigned by the ICCD. 34
These two values are both to be understood as identifiers of the resources to which they refer and
are therefore mapped with the CIDOC-CRM E42 Identifier class, which designates precisely the
identifiers, to be associated with the entity that describes the physical object (E22 Man-Made Object)
through the property P1 is identified by. Moreover, in order to distinguish the origin of the identifiers
of the two entities, the identifier of the entity has been associated with a type, expressed by the class
E55 Type, which is expressed respectively in “signature” and “unique asset code (NCT)”. The
33
OA: http://www.iccd.beniculturali.it/it/ricercanormative/29/oa-opere-oggetti-d-arte-3_00
information object (E73 Information Object) is also associated with its own identifier, which is the title
assigned to the resources, expressed by the CIDOC-CRM class E35 Title. The title is in fact the most
immediate reference to the information content of the object described.
4.2. People and roles
Person entities are represented by the class E21 Person. If persons act in groups or within complex
organizations, the class E74 Group will be used instead. The class E21 Person refers to all persons who
exist or existed in the past and of whom reliable historical information can be obtained. The description
of the person is linked to anthroponyms with references to possible vocabularies, dates and places of
birth and death, membership of groups, possible identifiers, qualifications and roles. As far as birth and
death of the person are concerned, these are described by the classes E67 Birth and E69 Death, which
express respectively the event of birth and the event of death, to which it is possible to link the date and
place of the event.
Persons present identifiers in the datasets we treated, which could be used to associate the context
of different collections to the entity. For archives the EAD elements that contain the name of the
producing entity are persname, famname and corporateBody: these are associated with an attribute,
authfilenumber, whose value represents the identifier that allows the link to the authority files in the
EAC-CPF format. For museums the ICCD standard provides for a separate authority file (AUT) in
which all the information relating to the authors of the works described in the OA files is given, to which
it is linked via the author identification code.
An interesting problem in the modeling of people’s information concerns the definition of the role
played by the actors in an event. In fact, in a context of integration of datasets coming from different
institutions when (e.g.) an anthroponym is common to both, the same person may have covered a
different function or role during his involvement with different resources (i.e.: Francesco di Marco
Datini is a merchant and a banker when he deals with his business, a commissioner when he
commissions his portrait to an artist, a sender or a receiver of a letter, etc.). CIDOC-CRM ontology
provides the P14 carried out by property to define the person responsible for an event. However, using
only this property it is not possible to define the role of the author with respect to the event, so all roles
are modelled as qualifications of the same person without distinction of the role carried out with respect
to the different events in which he may have participated. It is therefore necessary to distinguish between
the different roles played by the people involved in the creation and production of the object. To do this
it is necessary to extend the CIDOC-CRM ontology with the CRM-PC schema, which introduces,
among others, the property P14.1 in the role of. Along with that the new entity PC14 carried out by
acts as a glue and makes it possible to define the role of an actor with reference to a specific event.
PC14 carried out by is defined together with the properties P01 has domain and P02 has range,
indicating respectively the domain (event) and the range (actor) of the entity, and P14.1 in the role of,
defining the role of the actor in the event.
4.3. Location and places
Another important connection element in the modeling is defined by the class E53 Place. This class
comprises physical extents in space, independent from temporal phenomena and matter. The instances
of Place are referred to by toponyms that effectively represent the different appellation of places through
time and linguistic variations. Moreover most events have a place associated through the property P7
took place at that indicates where the event happened. As far as it was possible to pinpoint the exact
location of a place coordinates (latitude and longitude) where associated to it with the P168 place is
defined by property, as to determine the position into the physical space of Earth
There is also the necessity to define the physical location of the resource. It requires, in addition to
resource-specific identifiers, the indication of the current geographical and administrative location.
According to the EAD standard, the institution or agency responsible for providing intellectual
access to the described materials is described through the repository element. The OA schema also
includes the PVC field to document the current location (geographical and archival) of the resource.
In CIDOC-CRM such locations are also described through the E53 Place class, the P54 has current
permanent location property that has been chosen to define the current permanent location of the
physical object. The need to specify the current location of the resource arises because, although the
repository providing intellectual access usually also has physical custody of the materials, this is not
always the case. For example, a repository may assume responsibility for long-term intellectual access
to electronic documents, but the actual electronic files or data systems may continue to reside in the
office where they were created and maintained, or they may be held in long-term custody by entities
that can provide the appropriate technical facilities for preservation.
4.4. Object Description
As far as the modeling of technical data is concerned, these are to be considered in relation to the
physical object as they describe tangible characteristics of the object itself. We consider technical data
associated with the physical object to be information on:
- Material: comprises the concepts of materials that compose the objects (e.g: paper, canvas,
wood, ecc). It is mapped through the class E57 Material and it is associated with the object with the
P45 consists of property.
- Technique: techniques used during the production of the object, described using the P32 used
general technique property associated with the instances of the E12 Production event.
- Dimensions: comprises the information about all dimensions associated with the object (e.g:
width, length, height but also monetary value). Dimensions are described as instances of the E54
Dimension class that can be specified with other information through related properties such as P90 has
value and P91 has unit.
- State of preservation: the state of preservation is the information about the condition of the
object at a certain time; if there are no time references, the information is considered to be relative to
the time the object was cataloged and it is described with the E3 Condition State class linked to the
physical object through the P44 has condition property.
- Features carried by the object: comprises all the features that can be found on the object (e.g:
inscriptions) that are mapped through the P56 bears feature property.
In the EAD standard, the physdesc (physical description) element identifies the texture and all
information about the physical characteristics of the material described. The information is given as free
text directly within it or, alternatively, is subdivided into the elements dimension, extent (consistency
of the material described), genreform (characteristics of the physical format), physfacet (information on
the external/physical appearance of the material). The EAD standard also allows encoding the
information used to describe the state of preservation of the document: the specific information is
provided within the phystech element.
In the OA schema, the MTC field includes both the material and the technique used during the
production of the work, while the MIS field concerns the measurements of the physical support of the
work. The STCC field describes whether the state of the object is ‘good’, ‘fair’, ‘mediocre’, ‘bad’ or
whether the data is not available. Given the nature of the information in the STC field and the phystech
element, the condition is understood to be at the time of cataloging, if no other specification is present.
4.5. Events
An event is defined as a set of delimited and coherent processes and interactions, which bring about
changes in cultural, social or physical systems, and which modify the state of the entity they refer to. In
general, an event is qualified as any defined process that results in a change of the state of the entity it
refers to. The effects of an event are not necessarily permanent and incontrovertible, but the event itself
must be able to be documented. In order to document the event, some trace of it must remain in the form
of the elements that participated in it. In general, an event is understood as an interaction between:
● One or more participants
● Venue of the event
● Temporal extremes
In fact, although an event may appear as having an “instantaneous” effect, in reality any material
process or interaction has a spatio-temporal extension (this is why events are referred to as temporally
defined entities or “temporal entities34”, while other entities, such as people etc., are considered as
persistent entities or “persistent entities35”). We describe below the events that have the widest use
within the modeling performed so far. 3536
4.5.1. Production
The CIDOC class E12 Production comprises activities that aim to create one or more new objects
(i.e.: characterized by spatial extension36) from other materials. Production can be understood as a
specialization of the activity of modification, where “modification” means a reworking of the same
object, which does not change in substance, while “production” implies the creation of something “new”
from the initial materials, which will be different from the final result of production. New means
something that does not resemble in substance and form the materials involved in its production, or
something that acquires a new meaning at the level of documentation in the light of a modification,
different from what it had before. 37
In the case of works of art, the creation of a work of art is the production of a new physical object,
with different characteristics and bearing different meanings than the materials it is made of. One or
more participants (author(s)) take part in the production of the work in a given place and time. On the
other hand, as far as archival materials are concerned, production is to be understood as referring to the
object as a whole with its physical characteristics, made up of paper, wood, ink, etc., in order to create
a work of art. If we want to document the activity of writing the manuscript, therefore the conception
of the content, we refer to the class E65 Creation, which describes the creation event.
4.5.2. Creation
The class E65 Creation describes an event that leads to the creation of conceptual objects or
intangible products, such as legends, poems, texts, music, images, films, laws, etc.. Therefore, we chose
to associate the information content of the described material with the creation event, made by an author
(person, family or group), in a specific place and time.
The event “creation” has been associated with the writing of the documents of the Archivio di Stato
di (and can be used - in the same way - in similar contexts, linked to the manuscript heritage etc.), as it
was considered that the information content expressed by the text was to be considered separately from
the support on which it is reported. As a conceptual object the text therefore falls within the scope of
class E65 Creation.
Among the documents in the archive, it was established to map the creation event only for the register
document type, while the exchange letters event, specifically created for the modeling of these
resources, was chosen to be associated with the correspondence.
34
cfr.: http://www.cidoc-crm.org/Entity/E2-Temporal-Entity/Version-7.1.1
35
cfr.: http://www.cidoc-crm.org/Entity/E77-Persistent-Item/version-7.1.1
36
cfr.: René Descartes Principia Philosophiae Amsterdam 1644, LXIII.
4.5.3. Transfer of custody
The E10 Transfer of Custody class describes the transfer of physical custody or legal responsibility
for a physical object (so the reference is always to the E22 Man-Made Object entity). The use cases of
the entity E10 Transfer of Custody are:
● taking over custody (if no previous responsible person exists);
● end of custody (there is no subsequent keeper;
● transfer of custody from one person to another;
● taking over custody from an unknown subject (previous responsible subject is unknown, not
documented);
● loss of the work (the current custodian is unknown).
Note that the indication of the donor or the receiver of the custody is optional, as they may not be
documented. In addition, the legal responsibility may lie with a party separate from the physical
custodian(s) of the object, in which case the type of responsibility in place with respect to the transfer
of custody can be documented with the P2 has type property. In the case of physical custody, the object
must be physically in the possession of the keeper (in whole or in part in the case of compound objects).
In the OA schemas there is the field LA (other geographic-administrative locations), which basically
contains the changes of hand of the work documented over time: the beginning and end of the custody
at each institution is documented in a more or less precise way through the field PRD, which contains
the information on the time in which a work is documented at a particular institution. The relationship
between these places has been mapped with CIDOC through instances of E10 Transfer of Custody: the
chronology of the different passages can be reconstructed and thus the actual passage of custody can be
mapped37. If the temporal data had not been available, it would have been possible to use only the
property P49 has former or current keeper, which associates the physical object with its current and/or
previous custodians and which is also a “shortcut” to the more detailed path represented by the event
described by the class E10 Transfer of Custody. 38
Instead, it was decided to map the event in order to be able to keep track of the history expressed in
the OA tab. The last documented custodian of the work is actually reported in LC (current geo-
administrative location), which describes the current custodian of the work. So the last custody step is
between the last geo-administrative location reported in LA (based on the chronology) and the location
reported in LC. All the entities described are or have been keepers of the work. In order to express the
chronology (E52 Time-span), the PRDI field has been used (custody start date at the institution to which
it refers), to indicate the date on which the custody transfer event from one institution to another took
place.
4.5.4. Exchange of letters
The E9 Move event is used to describe changes to the physical position, i.e. the movement of the
physical object. The properties P27 moved from and P26 moved to connect to instances of the class E53
Place that describe - respectively - the starting and ending point of the move. This class, focusing on
the movement of the object, proved not to be the best solution for the representation of information
related to correspondence: a typology that involves the sending and receiving of a document and
therefore needs to express also essential information on the recipient and the sender. It was necessary,
therefore, to implement a new class, EL1 Exchange of letters, having as superclass E7 Activity, for the
definition of the exchange of letters. The two subclasses EL2 Send letter and EL3 Receive letter, indicate
more specifically the activity of sending and receiving a letter. By associating to these classes the
37
Also in this case, the modeling takes into account the need of domain experts to reconstruct the history (i.e. origin and provenance) of
documents, as part of the philological-historical investigation etc.
property P14 in the role of, it is possible to specify the actors involved and their respective roles of
participation (sender and receiver).
5. Browsing the data
As stated in the workflow section, once the data modeling based on the CIDOC-CRM ontology is
completed, data are converted in triples (TTL format) and uploaded to the Virtuoso triplestore. It is
therefore necessary here to describe how data access is performed. The methods implemented for
accessing the resources are:
● SPARQL endpoint (http://dev.restore.ovi.cnr.it:8890/sparql/): a service that lets the users write
sparql query to send their requests and obtain results as a response to the queries.
● Virtuoso Facet Browser (http://dev.restore.ovi.cnr.it:8890/fct/): a facility service for browsing
the resources in the Triplestore. It lets the users type keywords in a search bar to start
navigation.
● LodLive38 instance of RESTORE (http://dev.restore.ovi.cnr.it/lodlive/): it’s a visual
“navigator” of resources, a tool that enhances the RDF quality data with the effectiveness of
visualization through navigable graphs39
Those thus listed, are very versatile tools that allow in-depth semantic searches; however, they are
not really user-friendly. Therefore, having a user-friendly interface where data consultation and
browsing does not require knowledge of the sparql language becomes a must, and it should allow for
easy navigation, independent from the users’ web proficiency. In fact, the hypothesized target is very
heterogeneous in terms of digital literacy, ranging from citizens interested in research activities to the
experts in a research infrastructure. It is therefore necessary to have more or less specialized search
options, depending on the targeted user’s familiarity with surfing the net, and on the level of expertise
in the topic object of the webquest. For this reason, the platform offers different data visualization
solutions (in practice, each target segment must have its own specific offer). The team wants to stress
particularly on the definition and visualization of the relationships established between the several
entities described in the data.
In brief, the platform should make the data visible, modellable, savable and downloadable by all
users. All these operations must be made available within the platform.
RESTORE develops a user-friendly interface to let users browse the data regardless of their web
literacy. The interface is, nonetheless, one of the means, though the more intuitive and visually
appealing, by which the users interact with the data in the Virtuoso Triplestore. The interface simulates
the queries and allows the user to navigate the semantic relationships established between data using
CIDOC-CRM ontology. In fact, the aim of the platform is to have different ways of visualization and
browsing of information, so that the user can find the resource of interest and read the information on
the institution website that provided the information. For instance, a search made on “Francesco di
Marco Datini” gives records as results standing for letters, registries or person's information; the user
can scroll and refine the results and then, by selecting one among the entris, be linked to the institutions’
website. In fact, RESTORE doesn’t want to build a pure visualization tool of archival entries or artwork
entries, given that those are already made available by the partners’ interfaces. What the project achieves
is the integration of data in a semantic environment that will activate new ways to reveal information
through the relationships mapped during the previous stages of the work. In this perspective, the purpose
of the RESTORE interface is to make this browsing activity simple for all users, and to effectively link
the resource in a multifaceted context.
At the moment, an alpha version of the search interface was deployed and a testing session is set.
The search interface implemented consumes the data loaded into the Virtuoso triplestore in the form of
38
LodLive http://lodlive.it/
triples and simulates some simple sparql queries that can be performed on the sparql endpoint of the
project.
The triples have been loaded in the triplestore into distinct graphs (the distribution is based on the
original data providers) and are then modeled following the CIDOC-CRM.
First of all a search bar is made available in the search interface. The search bar is an intuitive
graphical element in which the user can type keywords to launch a search operation that goes through
all the data stored. The textual search by keywords corresponds to a query in which the results are
filtered on the basis of the keywords entered, searched through all the resources, without further
constraints. In order to refine the search results, two filters have been enabled:
- Collections: the user can select the collection to browse, narrowing the scope of the search. The
collections correspond to the graphs associated with the data in the triplestore. So, selecting a
collection is equivalent to defining the graph in a sparql query.
- Types of entity: the user can narrow the search by choosing the type of entity of interest. Almost
each entity in the triplestre has an associated type following the CIDOC ontology. So, choosing
a type equals specifying the type of entity to search in the sparql query.
The collection and the reference category are selected through the choice menu available within the
search bar. The user should type his request in the search bar and select the search options of interest,
if the search is successful, the resources corresponding to the selected terms will appear on the screen.
As for the appearance of the search results, a classic view in tabular form of the occurrences of what
is sought with relative information and links to the resource will appear after the search is launched.
Every record reports the label assigned to the resource, the name of the graph that contains it and the
type of entity; the information provided will also help the user to refine the search if needed. Moreover
a series of actions are enabled starting from the resource: it will be possible to i) access a specialized
view for the type of resource, ii) save the hyperlink of the resource, iii) save a citation and to access the
iv) lodlive view.
The triplified data allow the move from a traditional visualization of lists of data to their semantic
extensions with the relationships between resources highlighted. This mode provides for the exploration
of the graph through the construction of a series of paths which, starting from the entity sought, explores
other entities going through the relationships that bind them; it is always possible, nonetheless, to get
back to a simpler data navigation and to get to the original websites that provided them.
The team is implementing custom views for the different types of source, and specific tools are going
to be used according to the characteristics of the resources. The resources with a dedicated view will
be: Persons, Places and Documents. For letters, intended as a type of a document, a custom view has
been created, showing the information (metadata) of the letter, graphically placed next to the
transcription of the same letter, so that the user can check the information while reading. Moreover, a
customized view is being implemented with a specialized software: EVT (Edition Visualization
Technology) is an advanced software tool for digital scholarly editions. EVT was chosen because it
respects the following requirements: it is a lightweight, open source tool specifically designed to create
digital editions from encoded texts in the XML format, allowing the user to browse, explore and study
the digital editions through a user-friendly interface. Currently, EVT 3, the latest version to be released
in January 2022, is being used for current development. (in nota: EVT is distributed as open source
software, all current code base is available on GitHub https://github.com/evt-project/evt-viewer-
angular).
6. Conclusions and future works
We are planning to extend the platform with other elements and ways of visualization of the data. A
study on toponyms was predisposed and each toponym is being associated with the corresponding
places and its coordinates. As the study goes on we are deploying maps to investigate the movements
of the entities described in the Triplestore.
We are also extending the study of anthroponyms with the help of domain experts in order to have
a comprehensive list of the name variants of the same person. This will facilitate further searches on
actors and enhance the definition of relationships.
When the studies on these elements are concluded it will be a great help for the researchers and we’ll
be able to create new ways of visualization that will be appealing also for the general public because it
will be possible to enhance the section dedicated to storytelling with new functions.
On another note RESTORE now has a complete set of tools for transformation of data from the
original format to triples to visualization. The workflow can be replicated and applied to other resources
in other contexts.
7. Acknowledgements
Special thanks are due to the project partners: Archivio di Stato di Prato, the Museo di Palazzo
Pretorio and the Istituto Opera del Vocabolario Italiano. These institutions, in addition to the support
provided for the interpretation and correct management of the data, were invited to test the platform for
consulting the data provided within the RESTORE Project.
This evaluation process allowed the team to receive feedback from the data providers on the
implemented platform and in particular on the accuracy of the search keys and the results obtained.
We would also like to thank, for their collaboration and for clarifications on the best solutions to
proceed with the modeling, our colleagues Athina Kritsotaki and Eleni Tsoulouha from FORTH
(Foundation for Research and Technology, Greece), active with part of the RESTORE Team within the
SSHOC project. The Team works within the framework of Task 9.4 dealing with the Digital Humanities
and Heritage Science Communities Data Pilot, in which the solution mentioned here was also
developed.
8. References
[1] Wilkinson, M.D. et al. (2016). The FAIR Guiding Principles for scientific data management and
stewardship. Sci Data. 3, 160018. https://doi.org/10.1038/sdata.2016.18
[2] Wisser, K. M. (2011). Describing entities and identities: the development and structure of encoded
archival context—corporate bodies, persons, and families. Journal of library metadata,
[3] 11(3-4), 166-175. https://doi.org/10.1080/19386389.2011.629960
[4] Pitti, D. V. (2013). Enhancing Access to Contextual Information on Individuals, Families, and
Corporate Bodies for Archival Collections. http://dx.doi.org/10.17613/M61T0Q
[5] Ministero dei beni e delle attività culturali e del turismo. Istituto per il Catalogo centrale e la
documentazione (2018). Normativa OA – Opere e oggetti d’arte, Versione 3.00. Struttura dei dati
delle schede di catalogo ICCD. http://www.iccd.beniculturali.it/getFile.php?id=7508
[6] TEI Consortium, eds. TEI P5: Guidelines for Electronic Text Encoding and Interchange.
[7] Version 4.3.0. Last updated on 31st August 2021. TEI Consortium. http://www.tei-
c.org/Guidelines/P5/ (ultimo accesso 20/09/2021)