propose (with ML methods) probable technical terms (\trefi candidates)
I have the feeling that it should be possible to automatically classify text fragments as "probable technical terms" that could (should) be defined elsewhere.
This would be valuable in the SMGloM context to suggest additional math vocabulary. If successful, we could have a tool like the treifier (#1 (closed)) that helps build up the library in a more machine-oriented way.
Edited by Michael Kohlhase