【389】Implement N-grams using NLTK

Ref: Natural Language Toolkit

Ref: n-grams in python, four, five, six grams?

Ref: "Elegant n-gram generation in Python"

import nltk

sentence = """At eight o'clock on Thursday morning
Arthur didn't feel very good."""

# 1 gram

tokens = nltk.word_tokenize(sentence)

print("1 gram:\n", tokens, "\n")

# 2 grams

n = 2

tokens_2 = nltk.ngrams(tokens, n)

print("2 grams:\n", [i for i in tokens_2], "\n")

# 3 grams

n = 3

tokens_3 = nltk.ngrams(tokens, n)

print("3 grams:\n", [i for i in tokens_3], "\n")

# 4 grams

n = 4

tokens_4 = nltk.ngrams(tokens, n)

print("4 grams:\n", [i for i in tokens_4], "\n")

outputs:
1 gram:
 ['At', 'eight', "o'clock", 'on', 'Thursday', 'morning', 'Arthur', 'did', "n't", 'feel', 'very', 'good', '.'] 

2 grams:
 [('At', 'eight'), ('eight', "o'clock"), ("o'clock", 'on'), ('on', 'Thursday'), ('Thursday', 'morning'), ('morning', 'Arthur'), ('Arthur', 'did'), ('did', "n't"), ("n't", 'feel'), ('feel', 'very'), ('very', 'good'), ('good', '.')] 

3 grams:
 [('At', 'eight', "o'clock"), ('eight', "o'clock", 'on'), ("o'clock", 'on', 'Thursday'), ('on', 'Thursday', 'morning'), ('Thursday', 'morning', 'Arthur'), ('morning', 'Arthur', 'did'), ('Arthur', 'did', "n't"), ('did', "n't", 'feel'), ("n't", 'feel', 'very'), ('feel', 'very', 'good'), ('very', 'good', '.')] 

4 grams:
 [('At', 'eight', "o'clock", 'on'), ('eight', "o'clock", 'on', 'Thursday'), ("o'clock", 'on', 'Thursday', 'morning'), ('on', 'Thursday', 'morning', 'Arthur'), ('Thursday', 'morning', 'Arthur', 'did'), ('morning', 'Arthur', 'did', "n't"), ('Arthur', 'did', "n't", 'feel'), ('did', "n't", 'feel', 'very'), ("n't", 'feel', 'very', 'good'), ('feel', 'very', 'good', '.')]

Another method to output:

import nltk

sentence = """At eight o'clock on Thursday morning
Arthur didn't feel very good."""

# 1 gram

tokens = nltk.word_tokenize(sentence)

print("1 gram:\n", tokens, "\n")

# 2 grams

n = 2

tokens_2 = nltk.ngrams(tokens, n)

print("2 grams:\n", [' '.join(list(i)) for i in tokens_2], "\n")

# 3 grams

n = 3

tokens_3 = nltk.ngrams(tokens, n)

print("3 grams:\n", [' '.join(list(i)) for i in tokens_3], "\n")

# 4 grams

n = 4

tokens_4 = nltk.ngrams(tokens, n)

print("4 grams:\n", [' '.join(list(i)) for i in tokens_4], "\n")

outputs:
1 gram:
 ['At', 'eight', "o'clock", 'on', 'Thursday', 'morning', 'Arthur', 'did', "n't", 'feel', 'very', 'good', '.'] 

2 grams:
 ['At eight', "eight o'clock", "o'clock on", 'on Thursday', 'Thursday morning', 'morning Arthur', 'Arthur did', "did n't", "n't feel", 'feel very', 'very good', 'good .'] 

3 grams:
 ["At eight o'clock", "eight o'clock on", "o'clock on Thursday", 'on Thursday morning', 'Thursday morning Arthur', 'morning Arthur did', "Arthur did n't", "did n't feel", "n't feel very", 'feel very good', 'very good .'] 

4 grams:
 ["At eight o'clock on", "eight o'clock on Thursday", "o'clock on Thursday morning", 'on Thursday morning Arthur', 'Thursday morning Arthur did', "morning Arthur did n't", "Arthur did n't feel", "did n't feel very", "n't feel very good", 'feel very good .']

获取一段文字中的大写字母开头的词组和单词

import nltk
from nltk.corpus import stopwords
a = "I am Alex Lee. I am from Denman Prospect and I love this place very much. We don't like apple. The big one is good."
tokens = nltk.word_tokenize(a)
caps = []
for i in range(1, 4):
    for eles in nltk.ngrams(tokens, i):
        length = len(list(eles))
        for j in range(length):
            if eles[j][0].islower() or not eles[j][0].isalpha():
                break
            elif j == length - 1:
                caps.append(' '.join(list(eles)))

caps = list(set(caps))
caps = [c for c in caps if c.lower() not in stopwords.words('english')]
print(caps)

outputs:
['Denman', 'Prospect', 'Alex Lee', 'Lee', 'Alex', 'Denman Prospect']