Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Chunking NP, VP and PP phrases in Java (CoreNLP)

Tags:

java

nlp

I'm using Stanford CoreNLP and I'm aware it doesn't support chunking of sentences. What I'm looking for is, given an input sentence, to have something like this as output:

 [NP He ] [VP reckons ] [NP the current account deficit ] [VP will narrow ] [PP to ] [NP only # 1.8 billion ] [PP in ] [NP September ] . 

I also know OpenNLP apparently supports this feature, but I already wrote quite a lot of code using CoreNLP and I would't like having to switch. So, what I'm looking for is either an external library that can do this for me or ideas about implementing this feature in the most simple way (references to publications, links, everything is welcome), starting from the parse tree. I don't need this to be as accurate as state of the art chunkers, at least for now, so I'm looking to implement this fast and maybe change it in the future if needed.

like image 623
The Coding Monk Avatar asked Mar 15 '15 11:03

The Coding Monk


1 Answers

In the end I found the Illinois chunker that does just what I needed.

like image 60
The Coding Monk Avatar answered Oct 16 '22 02:10

The Coding Monk