Application of Linked Open Data to the coding and dissemination of Spanish Civil War photographic archives
| Date | 08 October 2019 |
| DOI | https://doi.org/10.1108/JD-06-2019-0112 |
| Published date | 08 October 2019 |
| Pages | 67-95 |
| Author | Jesús Robledano-Arillo,Diego Navarro-Bonilla,Julio Cerdá-Díaz |
| Subject Matter | Library & information science,Records management & preservation,Document management,Classification & cataloguing,Information behaviour & retrieval,Collection building & management,Scholarly communications/publishing,Information & knowledge management,Information management & governance,Information management,Information & communications technology,Internet |
Application of Linked Open
Data to the coding and
dissemination of Spanish Civil
War photographic archives
Jesús Robledano-Arillo, Diego Navarro-Bonilla and Julio Cerdá-Díaz
Department of Library and Information Science,
Universidad Carlos III de Madrid, Madrid, Spain
Abstract
Purpose –The purpose of this paper is to present a conceptual model for coding and dissemination of data
associated with historical photographic archives. The model is based on Linked Open Data technology and
seeks to exhaustively represent the most relevant characteristics for the tasks of contextualization of the
documentary groupings and units, management, document retrieval, dissemination and sharing of data about
the historical photographs.
Design/methodology/approach –An OWL ontology, called Ontophoto, was constructed following an
adaptation of the methodology proposed by Uschold and Gruninger and Gruninger and Fox. The ontology
was implemented using Protégé 5.5 software. Next a Graph DB® graph database application (Ontotext) was
created to generate a query system based on the SPARQL language. To validate the consistency and
effectiveness of the model and ontology, a competency questions methodology has been applied using a
sample from the Skogler photographic archive.
Findings –The model facilitates the generation of systems for dissemination and retrieval of iconographic
data for historical research, overcoming some of the limitations with respect to the design of methods of
content and contextual information representation for heritage photographic archives.
Research limitations/implications –This study is based on a sample. Future work should consider the
implementation of the model on the totality of a photographic collection.
Originality/value –This paper presents a comprehensive ontological model that allows the creation of
distributed systems of knowledge representation, which can be queried through SPARQL language.
Keywords Linked Open Data, Semantic Web, Ontologies, Digital humanities, OWL, Image retrieval,
Photographic archives, Press photography, Spanish Civil War
Paper type Research paper
Introduction
Information systems for photographic archives have been gaining relevance in recent
decades in light of the revaluation of photographs as essential sources for historical, social
and anthropological studies, as well as for illustrating and teaching history (Borchert,
1982; Pink, 2001; Achterberg, 2007; Riego, 1996; Riego and Vega, 1994; Pantoja Chaves,
2007, 2008; Rodríguez de las Heras, 2009; Sanchidrián Blanco, 2011; Díaz Barrado, 2012).
The transformations in the organization of documentary collections spurred by the digital
humanities are likewise a necessary context that informed this project. A growing number
of institutions that house photographic archives with historical valueare publicly offering
them via the internet in a multitude of proprietary digital catalogs and repositories, as well
as via social media services like Flickr and metadata collectors like Europeana. Despite the
great volume, iconographic research is hampered by the limitations of the traditional
search systems we might find in archives and libraries. Harris and Hepburn (2013) have
noted the difficulty historians face in finding suitable photographs for their research,
Journal of Documentation
Vol. 76 No. 1, 2020
pp. 67-95
© Emerald PublishingLimited
0022-0418
DOI 10.1108/JD-06-2019-0112
Received 10 June 2019
Revised 27 August 2019
Accepted 28 August 2019
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0022-0418.htm
This workwas conducted using the Protégéresource, which is supportedby grant GM10331601 from the
NationalInstitute of General Medical Sciencesof the US National Institutesof Health. The authors would
like to thank the Fototeca de Huesca for the help given in the research on the Skogler photo archive.
67
Application of
Linked Open
Data
showing how the use of photographs in scientific literature within the history field has not
increased with the growing availability of digital images through libraries and social
media services.
Description of the images in library and archive query systems has been based on
textual representation of the content. Many descriptive methods applied to photographic
archives in these types of systems, such as library and archive cataloging standards, have
not sufficiently considered the peculiarities of photography, as they were not intended
for this type of media (Benson, 2009). To this we can add lack of uniformity and
differences of opinion regarding coding forms, standardization of terminology, andquery
and navigation interfaces. This problem is exacerbated by the incomplete nature of the
documentary descriptions of the photographs, which frequently fail to examine their
iconographic representation or allow access to them using data on the thematic context of
production. Automated systems for image analysis and retrieval cannot be applied to
these types of archives to achieve complete and exhaustive descriptions of all the levels
necessary for content representation. Enser et al. (2007) have noted the impossibility of
dealing with all the analysis and description levels of the content of photographic images
by means of fully automated systems that operate on the pixels of the images. They base
their position not only on the difficulty of associating specific concepts with the graphic
patterns (which vary from image to image) that represent them in terms of color, texture,
shapes or compositional structures but also on the fact that much of the important
meaning of the image is not expressed through any graphic pattern. That is, many of the
image’s meanings have no visual presence whatsoever, or their visual incidence can only
be inferred through high-level interpretation processes available only to a specialized
human observer, such as the representation of time, the concrete representation of space,
specific events, abstract concepts, symbolic meanings and values. We are obliged,
therefore, to continue using text and human analysts to represent many of the concepts
and relationships that we can identify in the content of the images and other data
necessary for their contextualization and management.
The question being considered is how to achieve textual representation methods that
make it possible to support query systems for graphic content that are powerful enough to
support the activity of iconographic research. Queries in image retrieval systems for which
content representations are insufficiently systematized generate serious problems of
documentary silence and noise. These problems translate into efficiency and effectiveness
limitations in search and investigation processes based on historical photographic images.
For this reason, information systems are required that not only enable precise and
exhaustive access to historical photographic images but also perform iconographic
investigation processes effectively. Our starting hypothesis is the need to use ontologies
within the Linked Open Data (LOD) context of the Semantic Web (SW) for representing and
utilizing iconographic information. The power of these methods and technologies in
achieving greater efficiency and effectiveness in the representation and sharing of the
knowledge with the necessary levels of precision and exhaustiveness, as well as their great
ease and flexibility in feeding the diverse formalized knowledge sources available on the
Web, make them relevant solutions.
SW technology improves the effectiveness of information retrieval processes, making it
possible to reduce the search time and cost for retrieving the required information by
enabling software agents to automatically determine the resources, interpret them and
integrate them to provide improved information searching and retrieval (Wang and
Barnaghi, 2007). Moreover, SW helps with dissemination and sharing of knowledge of the
documentary archives by establishing a representation of universal knowledge through a
standardized modeling language (RDF), enables the creation of reusable data models
(through the RDF, RDFS and OWL languages), facilitates content querying through a
68
JD
76,1
To continue reading
Request your trial