NLP | Verb Forms Correction

Let’s understand this with an example :

1. Is our child training enough?
2. Is are our child training enough?

The verb ‘is’ can only be used with singular nouns. For plural nouns we use ‘are’. This problem is very common in the real world and we can correct this mistake by creating verb correction mappings that are used depending on whether there’s plural or singular noun in the chunk.

Code #1 : Defining the verb correction mappings

# singular to plural mapping

plural_verb_forms = {
        ('is', 'VBZ'): ('are', 'VBP'),
        ('was', 'VBD'): ('were', 'VBD')
        }

# plural to singular mapping
singular_verb_forms = {
        ('are', 'VBP'): ('is', 'VBZ'),
        ('were', 'VBD'): ('was', 'VBD')
        }

We are searching the chunk for the position of the first tagged word using the first_chunk_index() method. This method had a parameter ‘pred’ that takes a (word, tag) tuple and returns True or False.

Code #2 : first_chunk_index()

def first_chunk_index(chunk, pred, start = 0, step = 1):

    l = len(chunk)
    end = l if step > 0 else -1

    for i in range(start, end, step):
        if pred(chunk[i]):
            return i

    return None

The predicate function in the code below returns True if the tag in the (word, tag) argument starts with a given tag prefix. Else, false.

Code #3 :

def tag_startswith(prefix):
    def f(wt):
        return wt[1].startswith(prefix)
    return f

Code #4 : Let’s correct the verb forms

from transforms import correct_verbs

print ("Corrected verb forms : \n",
       correct_verbs([('is', 'VBZ'), ('our', 'PRP$'),
                      ('children', 'NNS'), ('learning', 'VBG')]))

Output :

Corrected verb forms : 
[('are', 'VBP'), ('our', 'PRP$'), ('children', 'NNS'), ('learning',
'VBG')]