Τhe potential of Large Language Models (LLMs) in Knowledge Engineering and Ontology Development, promise or peril?

Oral Histories represent narrative recollections of people’s past experiences and are important primary source material for social, cultural and historical research, among other areas. Much interest currently exists in data-driven and algorithmic approaches to oral history research, at the individual collection level and also at scale. However, the discoverability of oral history collections is currently limited. There is no universally accepted metadata schema or standard available for the oral history domain. In 2022 the Oral History Association (OHA) released an extensive compendium of standard-agnostic metadata elements covering all stages of the oral History lifecycle. Through mappings to Dublin Core, we present a bespoke subset of the OHA compendium as a metadata schema that is optimised to represent digital oral history collections as structured data. Seeking to enhance the findability and interoperability of oral history, our schema promotes the FAIR (Findable, Accessible, Interoperable and Reusable) principles for data sharing and discovery and opens pathways to linked data publication. The research presented here draws on the Mixed Methods Digital oral history (MeDoraH) project which seeks to analyse oral history recollections of the discipline of Digital Humanities using digital methods.