Add dc:subject to synsets
Closed this issue · 4 comments
In en-word.net, each synset has an ILI and an attribute called dc:subject. Subjects correspond to the lexicographer files in Princeton wordnet. They are broad semantic categories. In GermaNet, they are called "Semantische Felder" (semantic fields).
I propose to add dc:subject to OdeNet. It can easily be done by matching OdeNet with en-word.net via ILI. IMO it would be a great help.
One use case is when checking ILIs. Detecting a wrong ILI is one thing, finding and assigning a correct new ILI might be difficult or even impossible. The first step should be to make sure that a synset is assigned to the appropriate dc:subject. And it would be more helpful than just deleting an incorrect ILI.
The Excel file attached contains a table with columns ILI, subject and more.
It is the result of a join between the following datasets:
- OdeNet 1.3 (December 2020) (converted to RDF format using http://server1.nlp.insight-centre.org/gwn-converter/#)
- File Ili-map-wn30.ttl from https://github.com/globalwordnet/cili.
- English Wordnet 2020 (converted to RDF).
Attachment:
ILI mapping enriched with dc_subject OdeNet 1.3.xlsx
Together with Johann Bergh, we are currently automatically correcting ilis. If this is finished, it will be a good idea to add subjects as well. Give me a moment, please.
Thanks for the feedback and good luck with the ILIs!
Done.