Hands-On Python Deep Learning for the Web

Creating TF-IDF vectorizers of users and products

We will now create two different vectorizers one is for users and the other for products. We will need these vectorizers in place to determine the similarity between the requirements of the users and what the reviews tell us about any given product. First, we will create the vectorizer for users and display its shape:

user_vectorizer = TfidfVectorizer(tokenizer = WordPunctTokenizer().tokenize, max_features=1000)user_vectors = user_vectorizer.fit_transform(user_df['Text'])user_vectors.shape

Then, we will create the vectorizer for products:

product_vectorizer = TfidfVectorizer(tokenizer = WordPunctTokenizer().tokenize, max_features=1000)product_vectors = product_vectorizer.fit_transform(product_df['Text']) ...