I am studying Natural Language Processing. What could be the ways to detect negation? There are at least two forms of negation that I can think of.
- I do not like orange juice.
- I deny that I like orange juice.
Thanks.
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
You can image a task like NER. Your dataset will be a set of sentences with the "negation" word annotated and that's it.
E.g.
x = "I do not like orange juice"
y = [0, 0, 1, 0, 0, 0]
The model then is a classification one, and it depends on several factors which one would be the more appropriate.
