Toward Semantic Metadata Aggregation for DPLA and Beyond: A Report of the ALCTS CaMMS Heads of Cataloging Interest Group, Orlando, June 2016

Abstract

DPLA content hubs and service hubs face similar challenges in aggregating metadata. These include quality assurance, reconciliation of terms, and conforming source data to the DPLA application profile.

An area receiving special attention is the clarification and mapping of rights statements. In some cases, there is no information in the record and it needs to be supplied. In others, there may be notes with vague or irregular wording, and these need to be mapped to a controlled vocabulary in order to be useful in discovery systems (e.g., through faceting and filtering). Rightsstatements.org is helping to make this possible by providing unambiguous statements backed up by persistent URIs.

For both the NYPL and the MDL, serving as a DPLA hub aligns with their institutional missions. By aggregating and enriching cultural heritage data from hub participants, they make their collections more discoverable on the Web and provide a valuable public service. And in order to provide additional value, both the NYPL and the MDL hubs are considering ways to push enhanced metadata (e.g., place names enriched with geographic coordinates) back to their original repositories.

Practitioners and managers of cataloging and metadata services have an important role to play in large-scale aggregation. They can ensure that when data sets from multiple sources are combined and normalized, that the underlying data semantics are preserved. Knowledge of resource description standards and controlled vocabularies continue to be highly valued, but must be applied at scale. An understanding of schema crosswalks continues to be important for aligning metadata with target applications. Metadata audits and index-based faceting can expose problems, while tools like Open Refine and Python can be used for programmatic remediation.