Author Gender Metadata Augmentation of HathiTrust Digital Library

摘要

Bibliographic metadata is essential for digital libraryrnresource description. Especially as the size and number ofrnbibliographic entities grows, high-quality metadata enablesrnricher forms of digital library access, search, and use.rnMetadata records can be enriched through automatedrntechniques. For example, a digital humanities scholar mightrnuse the gender of a set of authors during their literaturernanalysis. In this study, we undertook to enrich the metadatarndescription of a large-scale digital library, the HathiTrustrn(HT) digital library, specifically by determining the genderrnof authors of the public domain portion of the collection.rnThe results are stored to a separate Solr index accessiblernthrough the HathiTrust Research Center services. Thisrnstudy, which successfully resolved in 78.9% of the casesrnthe gender of authors in the HT public domain corpus,rnsuggests future research directions in capturing andrnrepresenting the provenance of the contributing sources tornenhance trust, and in machine learning to resolve thernremaining names.