TF-IDF Vectorizer: Turning Words into Meaningful Numbers

-

When dealing with text data, simply counting words (like CountVectorizer does) isn’t always enough. Some words like “the”, “is”, or “this” appear very frequently, but they don’t really help in understanding the meaning of the text. To solve this, we use TF-IDF (Term Frequency – Inverse Document Frequency).

What is TF-IDF?

TF-IDF is a method to convert text into numbers by giving more importance to rare but meaningful words and less importance to common words.

  • TF (Term Frequency): How often a word appears in a document.

  • IDF (Inverse Document Frequency): How rare a word is across all documents.

  • TF × IDF: Final weight for a word. Common words get low scores, rare words get high scores.

Example (Step by Step)

Suppose we have 3 sentences:

  1. "I love my phone"

  2. "This phone has a great camera"

  3. "I love this camera"

  • Word “phone” appears often → gets lower importance.

  • Word “great” appears once → gets higher importance.

This way, TF-IDF highlights important keywords in documents.

Why Use TF-IDF?

TF-IDF in Python

from sklearn.feature_extraction.text import TfidfVectorizer

# Sample documents
documents = [
    "I love my phone",
    "This phone has a great camera",
    "I love this camera"
]

# Initialize TF-IDF Vectorizer
vectorizer = TfidfVectorizer()

# Fit and transform documents
X = vectorizer.fit_transform(documents)

# Show vocabulary
print("Vocabulary:", vectorizer.get_feature_names_out())

# Show TF-IDF values
print("\nTF-IDF Matrix:\n", X.toarray())

Important Parameters of TF-IDF Vectorizer

1. max_features

  • Limits the number of words in vocabulary.

  • Example: max_features=5 keeps only top 5 words.

2. ngram_range

  • Considers multiple words together (n-grams).

  • Example: ngram_range=(1,2) includes single words and pairs like “great camera”.

3. stop_words

  • Removes common meaningless words.

  • Example: stop_words='english' removes words like “the”, “is”.

4. min_df & max_df

  • min_df: Ignore words that appear too rarely.

  • max_df: Ignore words that appear too frequently.

  • Example: min_df=2 keeps only words that appear in at least 2 documents.

5. use_idf

  • Whether to apply IDF or not.

  • Example: use_idf=False makes it similar to CountVectorizer but normalized.

6. norm

  • Normalizes row values (so that long sentences don’t get higher scores).

  • Options: 'l1', 'l2', or None.

7. smooth_idf

  • Prevents division by zero when a word appears only once.

  • Default: True.

Comparison Table: CountVectorizer vs TfidfVectorizer

Feature CountVectorizer TfidfVectorizer
Output Word counts Weighted scores
Handles common words No Yes (down-weights them)
Highlights unique words No Yes
Use case Simple models Search engines, classification

Real-Life Analogy

Imagine a classroom discussion:

  • Common words like “the”, “is” are spoken by everyone → less important.

  • Unique terms like “neural networks” or “camera quality” are rare but meaningful → more important.
    TF-IDF ensures we pay attention to these important terms.

Final Thoughts

TF-IDF is a powerful improvement over simple word counts. It’s widely used in:

If you’re working with text and want your model to focus on meaningful words rather than just frequent ones, TF-IDF Vectorizer is your go-to tool.

Comments

Post a Comment

Popular posts from this blog

Understanding Data Leakage in Machine Learning: Causes, Examples, and Prevention

-
When building machine learning models, achieving high accuracy on your training or validation data is exciting — but sometimes, that success can be misleading. One common trap that causes overly optimistic results is data leakage . In this blog, we will explore: What is data leakage? Why is it harmful? Real-world examples to understand leakage better How to prevent data leakage in your projects What is Data Leakage? Data leakage happens when your machine learning model has access to information during training that it wouldn’t realistically have when making predictions in the real world. This extra information “leaks” from the future or from the test set, causing the model to cheat. When leakage occurs, your model's performance metrics (like accuracy, R², or F1-score ) look great, but this performance won’t generalize to new, unseen data. Why is Data Leakage a Problem? Overfitting to training data : The model learns shortcuts rather than meaningful pattern...
Read more

🌳 Understanding Maximum Leaf Nodes in Decision Trees (Scikit-Learn)

-
When working with decision trees in machine learning , one of the most common questions is: 👉 How many leaf nodes can a decision tree have, given certain constraints? Let’s walk through a real example. 📌 The Question We initialize a decision tree as follows: from sklearn.tree import DecisionTreeClassifier clf = DecisionTreeClassifier( max_depth =4, random_state =42) We’re told: The dataset has 100 samples and 10 features . The tree is a perfectly balanced binary tree . No pruning is applied. No additional stopping criteria (like min_samples_split , min_samples_leaf , or max_leaf_nodes ). 👉 What is the maximum possible number of leaf nodes in the decision tree? 🌲 Key Concepts 1. Depth vs. Levels A binary tree with max_depth = d has at most d splits from the root to a leaf. Depth d = 4 means there can be 4 splits from the root node down to a leaf node. 2. Maximum Leaf Nodes Formula In a perfectly balanced binary tree : Number of leaf n...
Read more

Linear Regression with and without Intercept: Explained Simply

-
If you’re new to machine learning, linear regression is one of the easiest models to understand. It tries to draw a straight line (or a flat plane, if there are more features) that best fits the data. When we use scikit-learn’s LinearRegression , we can decide whether the line should include an intercept (a constant number added at the end of the equation) or not. The Example Code import numpy as np from sklearn.linear_model import LinearRegression X = np.array([[12, 13], [23, 24], [24, 25], [35, 36], [36, 37], [37, 38]]) # Create target values y y = np.dot(X, np.array([4, 5])) + 6 reg1 = LinearRegression(fit_intercept=True).fit(X, y) s1 = reg1.score(X, y) reg2 = LinearRegression(fit_intercept=False).fit(X, y) s2 = reg2.score(X, y) What’s Happening Here? X is our input data (two features per row). y is the output we want to predict. We made it using this formula: y = 4 ∗ x 0 + 5 ∗ x 1 + 6 y = 4 * x_0 + 5 * x_1 + 6 This formula clearly has an intercept = 6 . Case 1: Wi...
Read more