CODEWITHSUNDEEP
nlptext-miningdata-science
Stefano De La Burczymuchito
Stefano De La Burczymuchito

Reputation: 133

NLP - identifying which adjective describes which noun in a sentence

I need a method / algorithm for identifying which adjective is related to which noun in a sentence.

Sample input:

"The product itself is good however this company has a terrible service"

As an output i'd like to get something like:

[product, good]
[service, terrible]

Could you please point me some algorithms / libraries that would help with such task?

Upvotes: 1

Views: 1373

Answers (1)

Wasi Ahmad
Wasi Ahmad

Reputation: 37731

You can use Stanford dependency parser, also this relevant paper. You can check their online tool as well. For example, for your sentence, you can get the following from Stanford parser.

Your query

The product itself is good however this company has a terrible service.

Tagging

The/DT product/NN itself/PRP is/VBZ good/JJ however/RB this/DT company/NN has/VBZ a/DT terrible/JJ service/NN ./.

Parse

(ROOT
  (S
    (NP (DT The) (NN product))
    (ADVP (PRP itself))
    (VP (VBZ is)
      (ADJP (JJ good))
      (SBAR
        (WHADVP (RB however))
        (S
          (NP (DT this) (NN company))
          (VP (VBZ has)
            (NP (DT a) (JJ terrible) (NN service))))))
    (. .)))

Universal dependencies

det(product-2, The-1)
nsubj(good-5, product-2)
advmod(good-5, itself-3)
cop(good-5, is-4)
root(ROOT-0, good-5)
advmod(has-9, however-6)
det(company-8, this-7)
nsubj(has-9, company-8)
dep(good-5, has-9)
det(service-12, a-10)
amod(service-12, terrible-11)
dobj(has-9, service-12)

Upvotes: 3

Related Questions