Student Forum

反式集体(TMDC)是一种特殊的元数据group of nearly a hundred information professionals, developed best practices for the description and classification of trans and gender diverse information resources. The culmination of their work, Best Practices for Trans and Gender Diverse Metadata give precedence to transparent practices, cultural sensitivity, correct identification, explicit descriptions of transphobia, and regular assessment of trans-related content. The TMDC report emphasizes the importance of minimizing potential harm and protecting privacy in metadata creation. Overall, the report aims to enhance the representation and inclusion of trans and gender diverse communities in cultural heritage institutions. The author of this paper was the primary organizer for and a co-author of the TMDC report. In this paper, I briefly summarize the collective’s investigation and outline its recommendations regarding common metadata standards, including RDA (Resource Description and Access), Library of Congress Classification (LCC) and others. Following this, I offer initial considerations about how the TMDC’s principles and recommendations affect Dublin Core metadata (both the Core and the DCMI metadata types).

本文提出的应用开放有关metadata in order to improve discoverability, provenance research, and restitution efforts for Nazi-looted art. Multinational and national online registries, including the German Lost Art Database and lootedart.com, were created in response to the 1998 Washington Conference on Holocaust-Era Assets. The next step is to expand upon the work of these repositories by creating robust linked metadata for the objects described within them. Linked open data will support data integration across institutions, increase discoverability, and enable machine learning and artificial intelligence applications in the looted art space. Included is a proposal for a data model that is adapted to the specific, contextualized needs of looted art objects and built using elements from the Europeana Data Model, Dublin Core, and RDA. Using these widely adopted frameworks in conjunction with standardized vocabularies will increase interoperability and support ease of implementation. Employing a hypothetical example, we illustrate how a knowledge graph can help researchers to sift large amounts of data, thereby increasing the chances of matching objects from disparate data sources.

Since the 1940s, superhero comics have occupied a pervasive position within modern popular culture, from books to movie franchises and video games. Contemporary comic characters have multiple storylines, are rebooted every few years, and crossover frequently with other comic series. Without the necessary background knowledge, a new comic reader faces a high barrier to entry. For comic books to remain open, accessible, and to meaningfully engage an audience as diverse as its content, an interconnected and flexible metadata structure is critical. We believe that improved metadata standards and structures can make comic books accessible to a new readership with a diverse range of backgrounds and experiences. By implementing a knowledge graph and metadata structure, a vast collection of comics can be transformed into an accessible, intuitive database. To demonstrate this improved functionality, we created a sample knowledge graph for two comics in the DC Universe: Nightwing and Superman Son of Kal-El. We combined a fan created and maintained comic book database, The Grand Comic Database (comics.org), with an open-source, fan created, comic book ontology, Comic Meta (comicmeta.org), and applied bespoke customizations to highlight possi

Makerspaces and Fab Labs are growing in popularity and now have large repositories of unique and nontraditional resources. However, there is a noted gap in the literature addressing the long-term management of Makerspace resources. Based on the digital resources of one Makerspace [The Champaign- Urbana Community Fab Lab (CUCFL)] we created a metadata schema useful to both staff and patrons to manage these items. Our central goals were to organize, improve search and discovery, and allow for adaptability. The CUCFL needed the schema to allow for change, so we originally created a simple schema to serve as a blueprint for a refined final version. For the CUCFL, the final version is an updated metadata schema adapted to serve as the information architecture for a new website functioning as a staff and public accessible library of their resources. With this example, we hope to encourage more research and development in this area. This project was supervised by Professor Inkyung Choi at the University of Illinois Urbana-Champaign.

Time-aligned textual transcriptions and the relationships between them and other files to which they relate can have complex creation narratives. This paper/presentation addresses critical questions related to cataloging and record creation for these types of digital artifacts. The approach appeals to WEMI and the file creation workflow as a way to analyze transcriptions and annotations as independent works from their source material. Arguments are presented for an analysis where transcriptions are independent expressions if not also independent works. According the Dublin Core 1:1 principle this means that Dublin Core based metadata schemas should support these resources with a separate description record. The two records—source media and transcriptions—ought to be related but separate. Modern practices of language documentation can complicate the expression of relationships because single files may include multiple expressions or both annotations and translations via different contributors.

Collaborative metadata creation allows organizations to build learning object repositories more efficiently when content authors complete metadata records as they build new learning objects. However, even with guidelines in place, human-generated metadata may be flawed, incomplete, or inaccurate. In addition, metadata guidelines may not exist for the duplication of learning objects. The unsystematic duplication of objects without differential metadata overloads the CMS with unnecessary content and hampers the search function. This paper reviews one such issue within an educational services company. The unchecked growth of duplicate learning objects has led to the creation of over one hundred thousand activities with titles identical to one or more other activities. Metadata specialists are leading a project to manually review and disambiguate content to improve function and searchability. Recommendations for more conscientious metadata creation can be learned from this project and if employed, would prevent the need for this type of manual disambiguation in the future.