A RDF-based approach to metadata crosswalk for semantic interoperability at the data element level
| Date | 15 June 2015 |
| Published date | 15 June 2015 |
| Pages | 175-194 |
| DOI | https://doi.org/10.1108/LHT-08-2014-0078 |
| Author | Ya-Ning Chen |
| Subject Matter | Library & information science,Librarianship/library management,Library technology |
A RDF-based approach to
metadata crosswalk for semantic
interoperability at the data
element level
Ya-Ning Chen
Department of Information and Library Science, Tamkang University,
New Taipei City, Taiwan
Abstract
Purpose –The purpose of this paper is to propose a Resource Description Framework (RDF)-based
approach to transform metadata crosswalking from equivalent lexical element mapping into semantic
mapping with various contextual relationships. RDF is used as a crosswalk model to represent the
contextual relationships implicitly embedded between described objects and their elements, including
semantic,hierarchical, granular, syntacticand multiple object relationshipsto achieve semantic metadata
interoperabilityat the data element level.
Design/methodology/approach –This paper uses RDF to translate metadata elements and their
relationships into semantic expressions, and also as a data model to define the syntax for element
mapping. The feasibility of the proposed approach for semantic metadata crosswalking is examined
based on two use cases –the Archives of Navy Ships Project and the Digital Artifacts Project of
National Palace Museum in Taipei –both from the Taiwan e-Learning and Digital Archives Program.
Findings –As the model developed is based on RDF-based expressions, unsolved issues related to
crosswalking, such as sets of shared terms, and contextual relationships embedded between described
objects and theirmetadata elements could be manifestedinto a semantic representation. Corresponding
element mapping andmapping rules can be specified without ambiguity to achieve semantic metadata
interoperability.
Research limitations/implications –Five steps were developed to clarify the details of the
RDF-based crosswalk. The RDF-based expressions can also serve as a basis from which to develop linked
data and Semantic Web applications. More use cases including biodiversity artifacts of natural history
museums and literary works of libraries, and conditions, constraints and card inality of metadata data
elements will berequired to make revisions to fine tune the proposed RDF-based metadatacrosswalk.
Originality/value –In addition to reviving contextual relationships embedded between described
objects and their metadata elements, nine types of mapping rules were developed to achieve a semantic
metadata crosswalk which will facilitate the design of related mapping software. Furthermore, the
proposed approach complements existing crosswalking documents provided by authoritative
organizations, and enriches mapping language developed by the CIDOC community.
Keywords Digital libraries, Digital archives, RDF, Metadata, Crosswalk, Semantic interoperability
Paper type Research paper
Introduction
In digital libraries, metadata playsan essential role in resourcedescription and discovery.
Many existing generic and domain-specific metadata standards have been developed by
various user communities and adopted as schema to build up digital library systems or
Library Hi Tech
Vol. 33 No. 2, 2015
pp. 175-194
©Emerald Group Publis hing Limited
0737-8831
DOI 10.1108/LHT-08-2014-0078
Received 7 August 2014
Revised 20 March 2015
Accepted 5 April 2015
The current issue and full text archive of this journal is available on Emerald Insight at:
www.emeraldinsight.com/0737-8831.htm
The author is grateful for the helpful comments of the anonymous reviewers. Furthermore, the
author thanks the National Palace Museum (hereafter NPM) in Taipei for offering metadata
instance of Table IV selected from Digital Archiving System of NPM Antiquities and Dr Wei-Lee
Lu of ROC Naval Academy for provision of metadata instance listed at Table II.
175
A RDF-based
approach to
metadata
crosswalk
repositories. Each metadata standard has been developed for a different purpose to
serve the needs of a particular user community or domain (Chan and Zeng, 2006).
With the use of varying metadata standards, integration or sharing of heterogeneous
metadata betweendigital library systems orrepositories, i.e., the interoperability between
standards becomes an issue. Metadata interoperability can be divided into schema,
record and repository levels (Chan and Zeng, 2006). According to definition proposed
by Chan and Zeng (2006), “crosswalks are by far the most commonly used method to
enable interoperability between metadata schemas at the schema level.”Most current
crosswalks are a semantic mapping of elements without a metadata conversion
specification (St Pierre and LaPlant, 1998). This reveals that most metadata crosswalks
are still focussedon mapping equivalent lexical definitions of metadata elementsbetween
source and target metadata standards. However, in addition to lexical semantics,
a complete semantic crosswalk has to take into account other issues such as common
terminologies, granularity of elements, syntax, semantic and hierarchical relationships
embedded between elements, multiple objectsand their relationships, and corresponding
mapping rules (St Pierre and LaPlant, 1998).
Literature review
In computer science automatic schema and ontology matching approaches have
been developed for data integration, e-business, data warehousing and semantic query
processing. They can achieve approximate matching between source and target metadata
elements (Rahm, 2011; Rahm and Bernstein, 2001) or lexical matching without contextual
relationships (Kitamura et al., 2008). The reason lies in each metadata standard is rooted in
a distinct domain with heterogeneous semantics (Nogueras-Iso et al., 2005), such as
Categories for the Description of Works of Art (hereafter CDWA) from museums and
Encoded Archival Description (hereafter EAD) from archives. On the other hand, in recent
years numerous projects have been initiated to achieve crosswalks between different
metadata standards and their applications, including selected two or more metadata
standards to build up a mapping mechanism to share metadata elements between similar
or different communities (Cao et al., 2004; Chandrakar, 2002, 2005; Godby et al., 2008;
Lagunas and Basurto, 2005; Lightle and Ridgway, 2003; Shepherd and West, 2003).
The issues and practices explored by these studies are useful guidelines in practical
mapping for similar cases and projects, but current proposed approaches have focussed
on mapping equivalent semantic definitions between elements through by representation
with a table or chart. In addition to table or chart, Morform and XSL has been used to
specify the syntax (Godby et al., 2008), as well as a set of step-by-step procedures (Lim et al.,
2012) and a mapping language (Kondylakis et al., 2006) has been proposed for metadata
crosswalk. However, to this day no attempts have been made to develop a generic
approach to semantic metadata crosswalks at the data element level.
Semantic metadata crosswalks at the data element level are more than equivalent
mapping based on lexical definitions of metadata elements. Researchers (Chan and Zeng,
2006; Dunsire et al., 2011; Park, 2002) have regarded this kind of crosswalk as lexical
mapping that is based on lexical form, appearance or meanings of metadata elements.
Lexical crosswalks still require a set of common or shared terms to reconcile their
semantic heterogeneity (Cao et al., 2004; St Pierre and LaPlant, 1998). Most current lexical
crosswalking practices lack explicit contextual information and relationships
for mapping (Dunsire et al., 2011; St Pierre and LaPlant, 1998; Woodley, 2008).
Thus, semantic (St Pierre and LaPlant, 1998), hierarchical (Cao et al., 2004; Lim et al., 2012;
St Pierre and LaPlant, 1998; Woodley, 2008), granular (i.e. many-to-one and one-to-many)
176
LHT
33,2
To continue reading
Request your trial