Tfidf and a BOW on same data

In this exercise, you will transform the review column of the Amazon product reviews using both a bag-of-words and a tfidf transformation.

Build both vectorizers, specifying only the maximum number of features to be equal to 100. Create DataFrames after the transformation and print the top 5 rows of each.

Be careful how you specify the maximum number of features in the vocabulary. A large vocabulary size can result in your session being disconnected.

Este ejercicio forma parte del curso

Sentiment Analysis in Python

Ver curso

Instrucciones del ejercicio

Import the BOW and Tfidf vectorizers.
Build and fit a BOW and a Tfidf vectorizer from the review column and limit the number of created features to 100.
Create DataFrames from the transformed vector representations.

Ejercicio interactivo práctico

Prueba este ejercicio completando el código de muestra.

# Import the required packages
____

# Build a BOW and tfidf vectorizers from the review column and with max of 100 features
vect1 = ____(____=100).____(____.____)
vect2 = ____(____=100).____(____.____) 

# Transform the vectorizers
X1 = vect1.transform(reviews.review)
X2 = vect2.transform(reviews.review)
# Create DataFrames from the vectorizers 
X_df1 = pd.DataFrame(X1.____, columns=____.____)
X_df2 = pd.DataFrame(X2.____, columns=____.____)
print('Top 5 rows using BOW: \n', X_df1.head())
print('Top 5 rows using tfidf: \n', X_df2.head())

Editar y ejecutar código