bag_of_words.py

import numpy as np
from nltk import word_tokenize

corpus = ["Life is short but life can seem long","The night is near"]

#definir deux phrases du corpus
sentence_1 = "Life is short but life can seem long"
sentence_2 = "The night is near"

# fonction retournant un vocabulaire
def vocabulary(corpus):
    voc = []
    for sentence in corpus:
        words = word_tokenize(sentence.lower())
        voc.extend(words)
        
    voc_clean= []
    for w in voc: 
        if w not in voc_clean: 
            voc_clean.append(w)
    return voc_clean


# fonction retournant un sac de mots 
def bagofwords(sentence,corpus):
    vocab = vocabulary(corpus)
    sentence_words  = words = word_tokenize(sentence.lower())
    bag_of_words = np.zeros(len(vocab))
    for w_in_sentence in sentence_words :
        for i,w in enumerate(vocab) : 
            if w == w_in_sentence : 
                bag_of_words[i] += 1
    return bag_of_words

sent1.py

print(bagofwords(sentence_1,corpus))

sent2.py

print(bagofwords(sentence_2,corpus))