Digitizing Metadata of a University Fashion Collection’s Holdings Using OCR and Costume Core

Abstract

To address the lack of standardized metadata in the historic dress and costume field and increase the searchability and discoverability of online fashion collections, a university collection adopted Costume Core, a metadata schema with controlled vocabulary for fashion artifacts, to develop the digital collection and to assess and improve the usability of the schema. This paper presents a case study delving into the process, benefits, and limitations of using Costume Core to catalog fashion artifacts. It includes a discussion on converting handwritten and typewritten metadata from physical catalog cards into a digital text format using Optical Character Recognition (OCR) technology. Costume Core provided a fundamental framework for digitizing and cataloging the university fashion collection while ensuring consistency in the terminology used to describe artifacts in the collection. Nevertheless, there is room for improvement for Costume Core, particularly by expanding the schema by adding controlled terms to increase the accuracy of terms chosen to describe fashion artifacts. Therefore, work is underway to increase its usability, thus promoting its adoption among other fashion collections.