Content Analysis of CONTENTdm Item Level Records and Their Aggregated Metadata in WorldCat

Content Analysis of CONTENTdm Item Level Records and Their Aggregated Metadata in WorldCat

Iris Lee (American Museum of Natural History, USA) and Mary Tyson (Whitney Museum of American Art, USA)
DOI: 10.4018/978-1-4666-2991-2.ch003
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

This study examines the content in CONTENTdm item level records and their aggregated metadata in WorldCat. The research ascertains some amount of loss of contextualization in semantic meaning of digitized primary resources. Data was collected from non-probability sampling of CONTENTdm item records selected randomly from CONTENTdm’s Collection of Collections Web page. The same aggregated digitized resources were retrieved in WorldCat, and the data was recorded and analyzed against the CONTENTdm records. Evidence shows that the value of the metadata is altered or ambiguous when the local field name is divorced from the descriptive metadata. Contextual meaning is lost when metadata is shared. In fact, the study reveals that the de-contextualization problem is two-fold: some inconsistent mapping from aggregated metadata and lack of clarity in the metadata itself.
Chapter Preview
Top

Background And Literature Review

A full color giga-pixel photograph is a much richer visual experience than a high-contrast black and white copy. As Coyle (2005) says, “… metadata is not the world, it is how we see the world at some moment in time for some purpose” (p. 160). The full information in a record is a surrogate for the real thing, and as such, metadata is a surrogate for the real thing found online (Coyle, 2005). However, when at one time, a user would sit in the library and look up resources in the ILS, now many kinds of users, from many locations, with many different purposes, from people to crawlers, all access metadata through the World Wide Web.

Research on metadata spans interoperability to the implementation of standards with quality and shareability at the center. Howe (2011) said about quality that to have good metadata you must make good metadata meaning automatically generated metadata will not suffice. Metadata that is shareable means it can be mapped to schemas such as DC and others. Riley and Shepherd (2009), as well as Suleman (2001), describe shareable as a known concept distinct from local. It is with this shareable structure that metadata can be mashed up outside of relational mapping (included in commercial sources such as Google, Wikipedia, the Atlas Initiative).

Complete Chapter List

Search this Book:
Reset