NLP Academy · 课时

使用 scikit-learn 实现 TF-IDF

用几行代码将语料库向量化

使用 scikit-learn 实现 TF-IDF 是 CoddyKit 上的免费 NLP Academy 课时。这是第 3 节课，共 4 节。你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。这是 NLP Academy 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 NLP Academy 课程共包含 4 节课。

本课时的部分内容尚未翻译，以英文显示。

No Need to Hand-Code It

You understand the math, so now let scikit-learn do the heavy lifting. Its TfidfVectorizer turns raw documents into a weighted matrix in a few lines.

Import the Vectorizer

Everything lives in the feature_extraction.text module. Import TfidfVectorizer and you are ready to vectorize any list of text strings.

from sklearn.feature_extraction.text import TfidfVectorizer

Your Corpus Is a List

A corpus is just a Python list of strings, one per document. Each entry is the full text you want scored and compared.

corpus = [
    "the cat sat on the mat",
    "the dog chased the cat",
]

Fit and Transform

Call fit_transform to learn the vocabulary and compute TF-IDF in one step. It returns a sparse matrix of weighted features.

vec = TfidfVectorizer()
X = vec.fit_transform(corpus)
print(X.shape)

What fit Learned

The fit step builds the vocabulary and the IDF values from your corpus. After this the vectorizer knows every term and how rare it is.

Inspect the Vocabulary

You can list the learned feature names to see the columns. get_feature_names_out shows each word in vocabulary order. 🔎

print(vec.get_feature_names_out())

The Output Is Sparse

Most documents use only a few words, so the matrix is mostly zeros. scikit-learn stores it as a memory-saving sparse matrix by default.

Peek at Real Numbers

Convert a row to a dense array to actually read the weights. The fillers near zero and topic words stand out clearly.

print(X.toarray()[0].round(3))

Tune With Parameters

Handy options let you drop rare or common terms instantly. Set min_df and stop_words to clean the vocabulary as you vectorize.

vec = TfidfVectorizer(stop_words="english", min_df=2)

Reuse on New Text

Fit once on training data, then call transform on fresh documents. New text is mapped into the exact same vocabulary and IDF scale.

new_docs = ["a new cat appeared"]
X_new = vec.transform(new_docs)

Ready for a Model

This weighted matrix plugs straight into any scikit-learn classifier. TF-IDF features are a strong, fast baseline for real text tasks.

Quick Check

Which method learns the vocabulary and computes the TF-IDF matrix together?

Recap

You imported TfidfVectorizer, fit it on a corpus, inspected the sparse output, and learned to reuse it on new text. The math is now a one-liner. ✅

常见问题解答

「使用 scikit-learn 实现 TF-IDF」课时是免费的吗？

是的 — 「使用 scikit-learn 实现 TF-IDF」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 NLP Academy 课程的其余内容，请升级到 CoddyKit PRO。 NLP Academy 课程共包含 4 节课。

「使用 scikit-learn 实现 TF-IDF」这节课中我会学到什么？

用几行代码将语料库向量化你通过在浏览器中直接运行的动手代码来练习 NLP Academy，全天候 AI 导师会在你学习这节课的过程中回答你的问题。

学习 NLP Academy 需要有经验吗？

无需任何先前经验。CoddyKit 上的 NLP Academy 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。这是第 3 节课，共 4 节。

「使用 scikit-learn 实现 TF-IDF」课时需要多长时间？

大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。

我能在这节 NLP Academy 课中编写并运行代码吗？

能。每节 NLP Academy 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。

此课程中的所有课时

← 返回 NLP Academy