Alignments / Crosswalks: Why

Why Create Alignments or Crosswalks?
Value of Aligning Metadata Vocabularies for Linked Data: In a Nutshell

The LOD movement aims to connect related data formerly isolated in repositories (often called “silos 1�7) and not previously linked. Cultural heritage institutions generate valuable data through tools such as finding aids, cataloging records, and databases, but this information is often sequestered inside archival information systems that cannot be easily linked to other data sources.

Linked data technologies and LOD resources can enable LAMs to...

 1�7 Enhance their already existing cataloging records, finding aids and other services
 1�7 Help users obtain new information and knowledge relevant to materials in their institution
 1�7 Make their collections available to users of external databases.


Value of Aligning Metadata Vocabularies for Linked Data: In More Detail

Linked Data (LD) offers great potential to connect library users to the vast array of information sources found outside the library’s virtual walls. By transforming library records into semantic statements such as RDF (Resource Description Framework) subject-predicate-object triples and mapping those nuggets of information to semantically-defined statements found in other data stores, libraries will offer users the chance to seamlessly integrate information from multiple sources. Using LD technologies, libraries can aggregate data based on the pieces/chunks of information they need from a dataset without integrating a whole database with their catalog or converting full metadata records. They can mash up metadata statements, rather than whole records, from a variety of datasets, based on the aligned metadata elements. Figure 1 presents such an example of bibliographic-data-centered mash up generated on-the-fly via the OpenAGRIS portal produced by the Food and Agriculture Organization of the United Nations.


Figure 1. Explanation of a bibliographic-data-centered mash-up generated on-the-fly at the OpenAGRIS portal Source: Based on OpenAGRIS (2012) http://agris.fao.org/openagris/search.do?recordID=PL2009000495.

In the center and upper-left of Figure 1 are the details of a bibliographic entry we normally see. On the two sides are various types of data from related resources such as Fish species, threatened (data from World Bank, middle-left), Thunnus obesus distribution map (data from Global Biodiversity Information Facility, lower-left), related articles from Nature (upper-right), and entries from DBpedia (the Linked Data version of Wikipedia, lower-right), and Global Capture Production statistic (not shown on this image).

Although the example given in Figure 1 is not in music-related areas (because we have not found any similar product derived from bibliographic records for music), it demonstrates how available LD datasets can be reused to enrich a bibliographic entry when presented to a user. In such an environment users are not merely using bibliographic data to find and obtain a book or article, but they are also gaining access to related contextual information provided by other data sources. An online bibliographic catalog or portal is no longer the ending point of an information search process. Instead, it provides new starting points for exploring related data and information, leading to new knowledge.

Yet, in order to mine these datasets outside of the Library’s virtual walls and obtain/link to what might be useful, one has to understand the data structures and metadata elements (also known as “properties 1�7) used by those datasets and find the corresponding linkable elements in library bibliographic data (usually known as “fields 1�7 in bibliographic records). The project reported here is one of the efforts of the research team that aim to research and build a Metadata Vocabulary Junction (http://lod-lam.slis.kent.edu) through the alignment of metadata elements that are used by different communities and data providers within and beyond the library world. Music domain metadata vocabularies, models of musical works 1�7 lifecycle, available music LD datasets, alignments at the levels of ontological classes, properties, and instances, are among the focuses of one of the projects (Gracy, Zeng, Skirvin, 2013). In this paper, we concentrate on the study of MARC-related elements in terms of their linkability for music information sources and discuss the barriers on the journey to using and becoming LD. Other studies we conducted that are related to archives (Gracy, 2014), electronic thesis and dissertations, and visual resources and cultural objects will be reported in other papers.



This work was supported by a grant from IMLS. It is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. reserved.