Testing Assumptions—Does Enhancing Subject Terms Increase Use of Digital Library Content?

Abstract

In modern library systems, access to the digital content is heavily dependent on effective metadata. The University of Florida (UF) Digital Collections (UFDC) are an actively growing, open access, digital library comprising over 500,000 records. As with any large-scale digital library project, a well-known challenge is the varying quality and quantity of legacy metadata available for each title. Inconsistent metadata makes digitized materials harder to find. If users cannot find the content they are looking for, a great deal of human effort has been wasted and the investment in digital collections is not being realized. Subject terms can be one of the most efficient methods for accessing desired materials, and subject terms created from controlled vocabularies deliver the most consistent results. To date, applying and editing subject metadata has been a record-by-record, labor-intensive process, making the prospect of retrospective projects cost-prohibitive. The UF team is investigating the capacity of research library staff to implement a Machine Assisted Indexing (MAI) system to automate the process of selecting and applying subject terms, based on the use of a rule set combined with controlled vocabularies, to the metadata of a body of already digitized content. To execute the project, the Smathers Libraries team at UF is collaborating with Access Innovations (AI) consultants to implement a machine-assisted indexing system to mitigate the challenges discussed above.