unibz-core / ontolex2skos Goto Github PK

Considering multiple executions of Ontolex2SKOS, the software must generate the same ID for the same entities. This must be done for facilitating data manipulation and management.

This behavior is working correctly in the current implementation for all entities except one: the schemes. Currently every individual with rdf:type skos:ConceptScheme changes its ID in every execution of the software.

The annexed file Result diff t21.docx presents the comparison of the results of two different executions using the same input file.

The annexed file Result diff t22.docx presents the comparison of the results of two executions, where in the second one the data was incremented w.r.t. the first.

The software generated an unexpected unnamed individual in a test execution.

The test files are (txt extension included for making possible the file upload - please remove it, keeping only the ttl extension):

• Used Input: Men Ontolex.ttl.txt

• Expected result (UNESCO Thesaurus fragment): men - edited.ttl.txt

• Obtained result (Ontolex2SKOS output): thesaurus.ttl.txt

With the introduction of this the PreferredSense class, new constraints also should come to ensure its uniqueness. This hold for each “language” (in fact SKOS allows multiple prefLabel assuming they are tagged with different language tags (@en or @es or..)).