CODEWITHSUNDEEP
pythonnltk
Ryan Scheuer
Ryan Scheuer

Reputation: 13

Tagging noun adjunct or attributive noun using NLTK

I want to use NLTK to recognize the difference between "red dress" and "dress shoes". The part of speech is the same for both, and none of the taggers I have tried have worked.

I'm aware that POS tagger can only pick up on one token at a time, which poses some problems. I'm very new to NLTK, and I'm sure I'm just missing something very basic here.

Upvotes: 1

Views: 208

Answers (1)

ᴘᴀɴᴀʏɪᴏᴛɪs
ᴘᴀɴᴀʏɪᴏᴛɪs

Reputation: 7519

Try Spacy

pip install spacy

Here is a small example to parse and visualize "dress shoes":

import spacy
from spacy import displacy

nlp = spacy.load('en')
doc = nlp("dress shoes")
displacy.serve(doc, style='dep')

red dress dress shoes

The relation between red (adjective) to dress (noun) is an adjectival modifier (amod), while the relation between dress (noun) and shoes (noun) is a compound.

These can be accessed via token.dep_ (for dependencies) and token.pos_ (for part of speech). Documentation here

Upvotes: 2

Related Questions