mex-atividade-4.2.py

# 1. Bag of Words

from sklearn.feature_extraction.text import CountVectorizer

corpus = ["gato dormindo", "cachorro latindo", "gato e cachorro brincando"]

vectorizer = CountVectorizer()
X = vectorizer.fit_transform(corpus)

print("Vocabulário:", vectorizer.get_feature_names_out())
print("Matriz de contagem (Bag of Words):")
print(X.toarray())

# 2. TF-IDF

from sklearn.feature_extraction.text import TfidfVectorizer

vectorizer = TfidfVectorizer()
X = vectorizer.fit_transform(corpus)

print("Vocabulário:", vectorizer.get_feature_names_out())
print("Matriz TF-IDF:")
print(X.toarray())

# 3. Word2Vec (Skip-gram)

import gensim
from gensim.models import Word2Vec
from nltk.tokenize import word_tokenize
import nltk

nltk.download('punkt_tab')

# Pré-processamento simples
corpus = ["gato dormindo", "cachorro latindo", "gato e cachorro brincando"]
tokenized = [word_tokenize(frase.lower()) for frase in corpus]

# Treinando modelo Skip-gram
model = Word2Vec(sentences=tokenized, vector_size=10, window=2, sg=1, min_count=1)

# Vetores das palavras
for word in model.wv.index_to_key:
    print(f"Vetor para '{word}': {model.wv[word]}")