Steps for supporting a new vocabulary

Step 1

Step 2

Metadata coordinator reviews the vocabulary based on the information provided in the request form.

This information follows the recommendations outlined by the Europeana Tech Task Force on Enrichment and Evaluation (see chapter 2.4. about selecting targets for semantic enrichment), namely:

Step 3

After the initial review, Europeana Operations team is consulted.

Europeana reserves the right to turn down the request if the vocabulary is not fulfilling the specified criteria and is not suitable for Europeana dereferencing efforts. Data provider is informed about the decision and reasoning behind it.

Step 4

If vocabulary is fulfilling the criteria, metadata coordinator checks which RDF/XML data profile the vocabulary content negotiates to - we make a request to URI with the Accept header “application/rdf+xml” - and creates a crosswalk between the format and EDM. It is possible that the vocabulary supports more than one data profile (preferably, vocabulary is based on standard data models, like, for example, SKOS). Library of Congress Subject Headings (LCSH), for example, supports SKOS as well as MADS, as is evident from this response. It is under the discretion of the Metadata Coordinator to select the data format that is most suitable for transformation.

Step 5

Metadata coordinator creates the contextual mapping (i.e. crosswalk) that focuses on overlaps between selected data profile and EDM contextual class. This can be done in partnership with the data provider who submitted the request.

Step 6

Metadata coordinator uses the contextual mapping as a general guidance for creating XSL stylesheet with a template for transforming vocabulary to EDM.

Data providers should note that some fields in vocabulary might not have an equivalent in EDM, and vice versa. In addition, there might be fields Europeana will omit from EDM output if, for example, they contain duplicated or information that is redundant in the light of Europeana’s wider enrichment goals. Such fields might be:

• Narrower relationships defined by skos:narrower and skos:narrowMatch

• Concept scheme information defined by skos:inScheme

If language coverage of the vocabulary goes beyond the languages supported by Europeana, the XSL stylesheet will leave out any unsupported values.

Step 7

The Manager of Aggregation Systems team adds the XSL stylesheet to Metis and publishes it on GitHub. All supported vocabularies are also listed in this spreadsheet.

Example - Dereferencing of LCSH vocabulary

Because LCSH vocabulary is in line with Europeana’s criteria for supported vocabularies, contextual mapping between SKOS and EDM contextual class skos:Concept was created.

Based on the crosswalk, XSL stylesheet was written and it is now used for LCSH dereferencing.

For example, when you run the XSLT process on this record this is the output:

Once XSL stylesheet is added to Metis, the following happens during Enrichment step:

Metis finds dereferenceable URI in the data:

Using XSL stylesheet Metis converts entity to EDM structure and adds it to the record. The link now resolves within the record: