ppt课件-europeanav1.0wp3chartingthelandscapeforedmprototyping.pptVIP

Interoperability Aspects in Europeana Antoine Isaac aisaac@few.vu.nl Workshop on Research Metadata in Context 7./8. September 2010, Nijmegen europeana.eu: the mission Making European cultural heritage better (web-)accessible Federating (online) cultural collections across countries and domains Hundreds of institutions, millions of objects europeana.eu in practice We rely on aggregating from our providers: Metadata References to digital objects We have a portal End-user “show-case” We will strive to become a metadata distributor Allowing partners to get enriched (contextualized) data for their objects Allowing third-parties to deploy object access functions similar to Europeana’s, in their own services Current status – provider data Very heterogeneous: different communities, different institutions, different interests and means Descriptions of original objects and digital objects uses hundreds of vocabularies, e.g.: Libraries: “MARC-style” records Museums: very diverse, richest ones with event-based descriptions (CIDOC-CRM) Archives: “EAD-style” hierarchical finding aids Cross-field “container” formats: METS Current status – provider data Grain varies Quality varies Free keyword indexing Explicit or implicit use of controlled vocabularies Adhoc vs. more standard (DDC, AAT, etc.) Persistent identifier usage not widespread (National) Libraries are doing better Current Europeana metadata stream Europeana Semantic Elements for ingestion of descriptive metadata and pointers to digital objects Dublin Core fields + Europeana-specific one Providers do the mapping from their data to ESE Ingestion process: OAI-PMH, still often via files (Fielded) full-text search using SOLR/Lucene Limitations of ESE Simple “flat” format Loosing richer (structured) data OK for full-text indexing and search Not ok for all the rest (display, access to data, richer search) Variations of DC field usage across collections dc:coverage dc:rights Digression: talk about rights? Lots