In order to be able to expose our lexicographical data via RDF we need a spec for the mapping. This should be based on the specification written in T150785: [Task] write spec for technical data model for Wiktionary support.
I have started a draft of mapping of the Lexeme DataModel to RDF: https://www.mediawiki.org/wiki/Extension:WikibaseLexeme/RDF_mapping
I have left some comments for different possible mappings of each element of the data model. I have focused on matching when relevant the Ontolex proposal, but it's maybe not the best idea: https://www.w3.org/2016/05/ontolex/