Chunking parenthesis with NLTK

Question

I am working with NLTK and I am trying to chunk (AIM 20-40-60) from the following text:

text = for more information refer to the Business Reporting Policy (AIM 20-40-60)

Currently I am using the following chunk pattern grammar = r"CHUNK:{<NN.*><CD>}" which is able to perfectly capture the AIM 20-40-60 part.

Nevertheless I also want the parenthesis () to be part of the chunk as well and since I am relatively new to regular expressions and chunking, I don't know the exact regEX pattern for capturing the parenthesis.

Try `grammar = r"\(CHUNK:{}\)"` (or `grammar = r"CHUNK:{\(\)}"`) — Wiktor Stribiżew, Mar 31 '17 at 09:01
I tried both versions but unfortunately it doesn't work. But thanks anyways :) — LoniF, Mar 31 '17 at 09:05
Well, you got the gist. Add `\(` and `\)` somewhere inside those tokens. — Wiktor Stribiżew, Mar 31 '17 at 09:06
The chunker pattern matches the sequence of POS tags, not the surface string. So instead of the the literal `()`, you need to match whatever tag they are given by the POS tagger. — lenz, Mar 31 '17 at 09:59
@lenz hey man you are right. just solved the issue. this is the pattern I used `grammar = r"CHUNK:{<\(><\)>}"` — LoniF, Mar 31 '17 at 10:06
Could you make that an answer? It would remove this from the list of unanswered questions "[regex] answers:0". — Yunnosch, Apr 01 '17 at 10:14

score 0 · Answer 1 · answered Apr 04 '17 at 06:23

0

The solution is the folloing

grammar = r"CHUNK:{<\(><NNP><CD><\)>}"

answered Apr 04 '17 at 06:23

LoniF

97
1
1
7

Chunking parenthesis with NLTK

1 Answers1