Generate relationships between tables and business glossary terms

Go to the project ddw-eureka-automations.

On the Project details page, click the Launch Workspace button.

Create a new SPARQL Query by clicking +Add > SPARQL Query.

Add the following query. Replace yourorgname with the name of your organization.

PREFIX : <https://yourorgname.linked.data.world/d/generate-relationships/>
PREFIX fn: <http://data.world/function/functions#>
PREFIX prov: <http://www.w3.org/ns/prov#>
PREFIX dwec: <https://dwec.data.world/v0/>

construct {
    ?table_IRI  dwec:subject ?term_IRI.
}
where {
    ?btType rdfs:subClassOf* dwec:BusinessTerm .
    ?term_IRI a ?btType ;
        dct:title ?BizTerm ;
    .
    MINUS{?term_IRI prov:wasGeneratedBy dwec:Eureka}
    BIND(LCASE(?BizTerm) AS ?BizTerm1) # standardize to lower case
    BIND(REPLACE(?BizTerm1, "[^a-z0-9]", "") AS ?BizTerm_ClusterKey) # collapse whitespace and punctuation

    ?table_IRI a dwec:DatabaseTable ;
        dct:identifier ?table_ID ;
    .
    BIND(LCASE(?table_ID) AS ?table_ID1) # standardize to lower case
    BIND(REPLACE(?table_ID1, "[^a-z0-9]", "") AS ?table_ClusterKey) # collapse whitespace and punctuation

    ##StringMatching
    BIND(fn:jaroWinklerDistance(?BizTerm_ClusterKey, ?table_ClusterKey) AS ?jw)
    BIND(fn:levenshteinDistance(?BizTerm_ClusterKey, ?table_ClusterKey) AS ?lev)

    FILTER(?jw > .8)
}

Save the query with:

Execute the SPARQL query from the previous step.

From the Results page, select Download the results > Save to dataset.

Set file name. The query name will automatically appear. Just delete the suffix .rq

Save the file to the ddw-catalogs dataset of your organization. You must set the file to auto-sync every hour.

Browse to the Glossary of your organization.

Browse to the tables you used for creating a list of glossary terms.