The implementations enumerated in the previous section are not meant to replace existing systems, but rather extend functionality by building on MARC data with linked data to enhance discovery. We can implement these services in a modular fashion, supported by independent and de-coupled APIs and standard web technologies. The modular implementation of a BibCard API aims for system stability by utilizing loosely coupled data sources implemented through web-based APIs.
We can use two related strategies to further grow a hybrid linked metadata environment. First, we can take existing MARC data, enrich it, and convert and cluster it in BIBFRAME graphs. Second, we can produce linked data in native RDF editors such as the Sinopia linked data editor and the Wikidata platform. Our Share-VDE participation, in which the Penn Libraries periodically receive enriched MARC and BIBFRAME graphs that have been converted by Share-VDE, supports the first strategy. In the near-term, Share-VDE will continue to provide this enrichment, transformation, and clustering service to Penn Libraries. A hosted database of linked data, along with external Share-VDE APIs, complement this data conversion and enable the discovery enhancements described in the previous section.
In the longer term, the Penn Libraries will continue the exploration of describing resources in linked data format including both Wikidata and Sinopia. Sinopia development is ongoing and now includes conversion of some BIBFRAME descriptions to MARC. There is also development toward ILS integration, which may make it possible to bring MARC equivalents of RDF resources into our catalogs to support enhanced discovery. Native RDF resources may also be stored in BIBFRAME Data Nodes (BF Nodes), hosted by the Library of Congress (LC), OCLC, Share-VDE, and others. BF Nodes are BIBFRAME knowledge bases maintained by institutions and service providers for the library community. International discussion and standards development for BF Nodes is ongoing and will be a critical piece for our future work in BIBFRAME.