Extractive vs Abstractive Summaries
Picking sentences vs writing new ones.
Extractive vs Abstractive Summaries is a free NLP Academy lesson on CoddyKit — lesson 1 of 4. You can read the complete lesson below for free — then practise it hands-on in the browser with a built-in code editor and a 24/7 AI tutor. It is part of the NLP Academy learning path, one of 4 lessons in the course, and your progress syncs across the web and the CoddyKit app.
Two Ways to Summarize
Summarization shrinks long text into a short version that keeps the key points. There are two main strategies, and knowing both helps you pick the right tool. 📝
Extractive Summaries
An extractive summary copies the most important sentences straight out of the source. Nothing is reworded, so the output is always faithful to the original.
Abstractive Summaries
An abstractive summary writes brand-new sentences that paraphrase the meaning. It reads more naturally, much like a human would summarize a story.
Copy vs Rewrite
The core difference is simple: extractive methods select existing text, while abstractive methods generate fresh text. One picks, the other paraphrases.
Why Extractive Is Safe
Because it reuses original sentences, extractive summarization rarely invents facts. That makes it a safe choice for legal, medical, or financial text.
Why Abstractive Is Fluent
Abstractive models can merge ideas from several sentences into one smooth line. The result is more concise and readable than a stack of copied sentences.
The Risk of Hallucination
Because it generates new words, an abstractive model can sometimes state facts that were never in the source. This mistake is called a hallucination. ⚠️
A Simple Extractive Idea
A basic extractive approach scores each sentence, then keeps the top ones. Here we just take the first two sentences as a quick baseline.
text = "NLP is fun. It powers search. It also writes summaries."
sentences = text.split(". ")
summary = ". ".join(sentences[:2])
print(summary)Modern Models Are Abstractive
Today's transformer summarizers are mostly abstractive. They read the whole text, then write a fresh, human-sounding summary from scratch.
Picking an Approach
Need guaranteed faithfulness? Choose extractive. Need a fluent, natural read and can verify facts? Reach for abstractive instead.
Both Need Good Input
Either approach only works well on clean, well-structured text. Garbled input leads to confusing or misleading summaries every time.
Quick Check
Which approach writes brand-new sentences instead of copying them?
Recap
You learned the two summarization styles: extractive copies key sentences and stays faithful, while abstractive paraphrases for fluency but can hallucinate. 🎯
Frequently asked questions
Is the “Extractive vs Abstractive Summaries” lesson free?
Yes — the full text of “Extractive vs Abstractive Summaries” is free to read here on the web, and the NLP Academy course includes 4 lessons in total. To practise it interactively (a built-in code editor and a 24/7 AI tutor) and unlock the rest of the NLP Academy course, upgrade to CoddyKit PRO.
What will I learn in “Extractive vs Abstractive Summaries”?
Picking sentences vs writing new ones. You practise NLP Academy with hands-on code you run directly in the browser, and a 24/7 AI tutor answers your questions as you work through the lesson.
Do I need any experience to start NLP Academy?
No prior experience is required. NLP Academy on CoddyKit is structured for beginners through advanced learners; this is — lesson 1 of 4, so you can start here or from the beginning and move at your own pace.
How long does the “Extractive vs Abstractive Summaries” lesson take?
Most CoddyKit lessons take about 5–10 minutes. Each one is bite-sized and interactive, so you make steady progress and pick up exactly where you left off across the web and the app.
Can I write and run code in this NLP Academy lesson?
Yes. Every NLP Academy lesson includes a built-in code editor, so you write and run real code right in your browser and get instant AI feedback — no local setup required.
All lessons in this course
- Extractive vs Abstractive Summaries
- Summarizing With a Seq2Seq Model
- Machine Translation in Practice
- Scoring Generation With ROUGE and BLEU