Text Preprocessing in Python | Set 2

Text Preprocessing is one of the initial steps of Natural Language Processing (NLP) that involves cleaning and transforming raw data into suitable data for further processing. It enhances the quality of the text makes it easier to work and improves the performance of machine learning models.

In this article, we will look at some more advanced text preprocessing techniques.

Prerequisites

Before starting with this article, you need to go through the Text Preprocessing in Python | Set 1.

Also, refer to this article to learn more about Natural Language Processing - Introduction to NLP

We can see the basic preprocessing steps when working with textual data. We can use these techniques to gain more insights into the data that we have. Let's import the necessary libraries.

# import the necessary libraries
import nltk
nltk.download('punkt')
nltk.download('averaged_perceptron_tagger')
import string
import re

Part of Speech Tagging

The part of speech explains how a word is used in a sentence. In a sentence, a word can have different contexts and semantic meanings. The basic natural language processing models like bag-of-words fail to identify these relations between words. Hence, we use part of speech tagging to mark a word to its part of speech tag based on its context in the data. It is also used to extract relationships between words.

from nltk.tokenize import word_tokenize
from nltk import pos_tag
# convert text into word_tokens with their tags
def pos_tagging(text):
  word_tokens = word_tokenize(text)
  return pos_tag(word_tokens)
pos_tagging('You just gave me a scare')

Output:

[('You', 'PRP'),
 ('just', 'RB'),
 ('gave', 'VBD'),
 ('me', 'PRP'),
 ('a', 'DT'),
 ('scare', 'NN')]

In the given example, PRP stands for personal pronoun, RB for adverb, VBD for verb past tense, DT for determiner and NN for noun. We can get the details of all the part of speech tags using the Penn Treebank tagset.

# download the tagset 
nltk.download('tagsets')
# extract information about the tag
nltk.help.upenn_tagset('NN')

Output:

NN: noun, common, singular or mass
    common-carrier cabbage knuckle-duster Casino afghan shed thermostat
    investment slide humour falloff slick wind hyena override subhumanity
    machinist ...

Chunking

Chunking is the process of extracting phrases from unstructured text and more structure to it. It is also known as shallow parsing. It is done on top of Part of Speech tagging. It groups word into "chunks", mainly of noun phrases. Chunking is done using regular expressions.

from nltk.tokenize import word_tokenize 
from nltk import pos_tag
# define chunking function with text and regular
# expression representing grammar as parameter
def chunking(text, grammar):
    word_tokens = word_tokenize(text)
# label words with part of speech
    word_pos = pos_tag(word_tokens)
# create a chunk parser using grammar
    chunkParser = nltk.RegexpParser(grammar)
# test it on the list of word tokens with tagged pos
    tree = chunkParser.parse(word_pos)
    for subtree in tree.subtrees():
        print(subtree)
    
sentence = 'the little yellow bird is flying in the sky'
grammar = "NP: {<DT>?<JJ>*<NN>}"
chunking(sentence, grammar)

Output:

(S
  (NP the/DT little/JJ yellow/JJ bird/NN)
  is/VBZ
  flying/VBG
  in/IN
  (NP the/DT sky/NN))
(NP the/DT little/JJ yellow/JJ bird/NN)
(NP the/DT sky/NN)

In the given example, grammar, which is defined using a simple regular expression rule. This rule says that an NP (Noun Phrase) chunk should be formed whenever the chunker finds an optional determiner (DT) followed by any number of adjectives (JJ) and then a noun (NN). Libraries like spaCy and Textblob are more suited for chunking.

Example:

Input: 'the little yellow bird is flying in the sky'

Output: (S (NP the/DT little/JJ yellow/JJ bird/NN) is/VBZ flying/VBG in/IN (NP the/DT sky/NN)) (NP the/DT little/JJ yellow/JJ bird/NN) (NP the/DT sky/NN)

Named Entity Recognition

As we know Named Entity Recognition is used to extract information from unstructured text. It is used to classify entities present in a text into categories like a person, organization, event, places, etc. It gives us detailed knowledge about the text and the relationships between the different entities.

from nltk.tokenize import word_tokenize
from nltk import pos_tag, ne_chunk
def named_entity_recognition(text):
    # tokenize the text
    word_tokens = word_tokenize(text)

    # part of speech tagging of words
    word_pos = pos_tag(word_tokens)

    # tree of word entities
    print(ne_chunk(word_pos))

text = 'Bill works for GeeksforGeeks so he went to Delhi for a meetup.'
named_entity_recognition(text)

Example:

Input: 'Bill works for GeeksforGeeks so he went to Delhi for a meetup.'

Output: (S
(PERSON Bill/NNP)
works/VBZ
for/IN
(ORGANIZATION GeeksforGeeks/NNP)
so/RB
he/PRP
went/VBD
to/TO
(GPE Delhi/NNP)
for/IN
a/DT
meetup/NN
./.)

Conclusion

In conclusion, natural language processing (NLP) plays a pivotal role in bridging the gap between human communication and computer understanding. As this field progresses, we can anticipate further innovations that will reshape how we communicate with and leverage the capabilities of intelligent systems in our daily lives and professional endeavors.