Metadata makeover

Submitted by Jeff Edmunds for the Digital Access Team

Discovery is driven by metadata. As the discovery ecosystem evolves toward a globalized, linked open data environment, the Digital Access Team in Cataloging & Metadata Services is working to make the Libraries’ metadata as clean, consistent, and granular as possible. To this end, the DAT recently collaborated with John Shank, Head Librarian at Berks, and members of the Thun Library staff, to effect a Metadata Makeover.

The makeover consisted of extracting copies of all records with Berks holdings from The CAT, performing extensive quality assurance protocols to assess which records were candidates for improvement, matching the records identified to their WorldCat counterparts, and then overwriting the existing records with superior versions. Over 8,000 records were enhanced. Improvements include:

  • More authorized headings for personal names
  • Additional subject headings
  • Foreign subject headings
  • Vernacular script (i.e. Russian records now include Cyrillic transcriptions in addition to Romanized ones)
  • Faceted Application of Subject Terminology (FAST) subject headings—an evolution of subject terms toward a Linked Open Data environment
  • Extensive Summaries and Contents notes, which not only provide rich corpuses of keywords that can be searched in The CAT and LionSearch but also contextualize materials whose character and content might otherwise seem ambiguous to users

In performing the makeover, we learned incidentally

  • that Berks holds 8,428 titles not held anywhere else in the Libraries’ system
  • that nearly 200 of these titles are not held by any other institutions worldwide (based on holdings set in WorldCat)
  • that nearly 300 items held by Berks do not have holdings set in WorldCat, meaning that the world is largely unaware that the Libraries in fact owns the items (a glitch that will be automatically remedied as a result of the makeover)

Plans are to continue with Metadata Makeovers, carefully targeting Libraries collections based on rigorous reviews of metadata quality.