Search code examples
stanford-nlp

How should I figure out the POS tag of "last" in this sentence?

I'm trying to use Stanford CoreNLP to tag a sentence.

"How long does a soccer game last?"

It seems on CoreNLP demo the token "last" is tagged as JJ instead of VB. Is there a way to fix this?

Solution

  • The short answer is "no." CoreNLP provides part of speech tags at a certain high but not perfect accuracy, and it will occasionally make mistakes. Beyond tweaking the tags yourself, there's no easy automatic way to have its accuracy go up.

    The longer answer is that you can always re-train the POS tagger on a custom tagged corpus, and then performance will be better on that corpus. This, however, involves a fairly substantial annotation effort tagging a large corpus of text with part of speech tags.