Comparative Study and Expansion of Metadata Standards for Historic Fashion Collections


The objective of this poster is to enhance the metadata standards applied in historic fashion collections. This is accomplished by expanding the controlled vocabulary and metadata elements to encompass the Costume Core and rectify any inadequacies. To achieve this goal, several methods are employed, including the incorporation of new descriptive terms to enable the precise description of artifacts during the re-cataloging of a university fashion collection in Costume Core. Additionally, new descriptors are identified through a technique called word embeddings, which involves using pre-trained natural language processing models to extract data from a conceptual latent space. Finally, crowdsourcing through surveys is conducted to gather insights into the usage of metadata for describing dress artifacts. Additionally, the presentation provides a preview of the Model Output Confirmative Helper Application, which streamlines the review process. It also highlights the commonly used metadata standards in the historic fashion industry, sample metadata supplied by respondents, and partial potential metadata to be appended to the Costume Core. As a result of the project, the expanded Costume Core is more comprehensive in describing fashion collections. It can be widely adopted by the fashion industry, promoting consistent metadata and increasing metadata interoperability.

Costume Core, Metadata, Crowdsourcing, Natural language processing, Cataloging, Digitization, Data standards, Controlled vocabulary, Standardization, Textile, Apparel, Fashion, Digital curation