Natural language processing for replacement candidate

[Ron89]

When query thesaurus.com for word under cursor, the synonym candidates in query result would all be in their original form. This is not ideal since more often than not the words under cursor are not in such form. And users may have to manually change the form of the synonyms they just chose from the list. It sort of defeats the purpose of having a choosing function in the first place.

I am thinking of using Python's NLTK library to change the candidates' form automatically, but that could add more dependencies which could be intrusive or confusing to some users.

Thoughts?