background image
for the Descriptions of Works of Art were considered
and in some cases used for vocabulary related to
classification of works.[3] Artist names were standar-
dised against the Italian national database of artist
names and a minimum record description was agreed
to and based on the art historical standards at the
national level in each of the participating partner
To group like objects, scholars agreed on several
sets of keywords based on best practice. Controlled
vocabulary was agreed for almost every descriptive
aspect including provenance (history of ownership),
materials of composition, category, technique(s),
process, place and dates. In addition, administrative
information such as the name of the cataloguer, the
language of the initial entry, and metadata details
about associated images were identified.
The database design went through fourteen
iterations during the prototype phase. The final
design includes twelve unique entities each with a
set of descriptors and relationship definitions. The
four primary entities are Object, People, Place and
Project. Together these entities form the basis for a
single entry.
The artwork is described in terms of its physical
characteristics such as materials and techniques of
composition and markings; how it relates to the Este
Family, through a link to the family member that
commissioned or collected it, and through extensive
provenance information; who created it, including a
link between the object and artist and biographical
details about the artist and their connection to the
Este Court; the current physical location of the
artwork; and the relation-ship between the artefacts
and a select set of themes taken from the Este
family archives that link the works with each
other and with the castle. In the context of the
ECA project each theme has three components:
the name of the project, its philosophical or
elemental theme, and the family member with
whom it is associated. Two examples of thematic
clusters are included in table 2.
This brings us to the third and probably the most
important aspect of the project - language.The
common language of the project was Italian - not
surprising, given the subject of the virtual collection.
However, it also became clear that all languages re-
presented by partner institutions would be required
in order to meet the needs of the potential visitors
to the site. As a result, all cataloguing was contrib-
uted first in the native language of the cataloguer -
French, German, Czech or Italian, with two sub-
sequent records contributed in Italian and/or English.
After meeting these initial requirements, cataloguers
could either continue contributing records in any
other official project language or request that another
cataloguer transcribe the Italian language record.
The database was developed using ORACLE9 on an
IBM H80 (6 CPU 500 MHz, 3 GB RAM in cluster
HM).The programming language was PHP (4.2.0).
Two interfaces were developed. Scholars logged
into a private section of the project Website and used
an interface designed specifically for them that in-
cluded checks against controlled vocabulary, as well
as indicators for level of completeness and related
cataloguing activities.The second interface is the
Preparatory model ­ modelletto
Preparatory model ­ bozzetto
Placchetta (Plaquette)
Modello preparatorio ­ modelletto
Modello preparatorio ­ bozzetto
Basso rilievo
Alto rilievo
Socharský model ­ modello
Socharská skica ­ bozzetto
Vysoký reliéf
Table 1: A sample from the three-term hierarchy of object classification developed
to aid in grouping objects for cataloguing, search and retrieval.
The Gilded
Dressing Rooms
The Apartment of
the Mirror
The myth of
Greco Roman
Alfonso I
Alfonso II
Table 2: Project themes link specific rooms within the
castle and adjacent apartments with the artistic themes
found in them and the member of the Este family
associated with their development.
Digicult_THI5_JS_090104 09.01.2004 14:33 Uhr Seite 22