Extractive summarization
Extractive summarization produces a shorter representation of source content by retaining selected original words, phrases, sentences, or passages and omitting other material.
Starting Data Analytics with Generative AI and Python
Extractive summarization creates a summary by selecting and combining important sentences, terms, or other passages from the source text.
- How it is used
- Summarizes short texts by selecting informative source material for quick, readable access to main information.
- How deep it goes
- Explains selection, rule-based scoring, and a frequency workflow, but does not develop advanced model internals.
- Examples
- Uses Portuguese customer reviews, frequency-ranked words, selected sentences, and comparisons with ChatGPT summaries.
- What you can do after
- Apply it via pandas and NLTK to preprocess reviews, count word frequencies, and generate short source-grounded summaries.
Read how Starting Data Analytics with Generative AI and Python explains extractive summarization →
Retrieval Augmented Generation, The Foundational Ideas
Extractive pruning reduces context by deleting selected material without generating replacement text.
- How it is used
- Reduces retrieved context for faithful factual recall and citation-sensitive answers.
- How deep it goes
- Explains deletion-based selection, ratio tuning, force-keep terms, and failure modes, while contrasting pruning with synthesis and cache eviction.
- Examples
- Uses LongLLMLingua, AttentionRAG, RECOMP extractive, TF-IDF filtering, quoted clauses, and force-kept domain terminology as examples.
- What you can do after
- Apply it via compression-ratio evaluation, force-token constraints, and task-specific selection between extraction and synthesis.
Read how Retrieval Augmented Generation, The Foundational Ideas explains extractive summarization →
Natural Language Processing in Action, Second Edition
Extractive summarization creates a summary by selecting and arranging passages or sentences from the source document.
- How it is used
- Uses source wording for concise document summaries, news curation, and other information-retrieval workflows.
- How deep it goes
- Defines relevance-based selection and verbatim reuse, but does not detail ranking, ordering, or implementation procedures.
- Examples
- Uses code autocomplete, repeated phrases, long-document summarization, and automated news curation as examples.
- What you can do after
- Apply it via transformers to summarize long documents or curate news articles automatically.
Read how Natural Language Processing in Action, Second Edition explains extractive summarization →
Effective Conversational AI
Extractive summarization is a summarization technique that constructs a shorter representation by selecting or extracting information from source content.
- How it is used
- Supports structured handoff summaries by extracting important transcript details for downstream agents.
- How deep it goes
- Explains prompt instructions, examples, output typing, model parameters, and hallucination verification, but not extraction internals.
- Examples
- Uses claims conversations containing TaxID, MemberID, ClaimID, dates, provider details, and a request for human assistance.
- What you can do after
- Apply it via an LLM by specifying fields, demonstrating desired output, tuning parameters, and checking extracted values against source text.
Read how Effective Conversational AI explains extractive summarization →
Building LLM Applications with DSPy
An extractive summary is a shortened text composed of passages selected from a source.
- How it is used
- Distinguishes summaries assembled from source passages from summaries phrased anew.
- How deep it goes
- Defines passage selection and discusses coherence, relevance, and readability, but gives no extraction algorithm.
- What you can do after
- Understand how extractive summaries preserve source wording and why selected passages may read poorly.
Read how Building LLM Applications with DSPy explains extractive summarization →
Also covered in
What to know first
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- LLM-as-a-judgeLLM-as-a-judge is an evaluation technique in which a language model assesses or compares outputs according to specified criteria.
- embeddingAn embedding is a numerical vector representation of an item in a feature space, encoding selected properties or relationships for computational tasks such as comparison or retrieval.
Related concepts
- summarizationkind of · Summarization is the production of a shorter representation of source information that preserves its important content or meaning.
- abstractive summarizationcontrasts with · Abstractive summarization generates a concise account of source information using newly composed wording rather than only selecting text from the source.
- generationkind of · Generation is the production of an output by a computational model from supplied input or context, often using patterns learned from data.
- fusion-in-decodercontrasts with · Fusion-in-decoder is a retrieval-generation method that independently encodes multiple passages and fuses their representations in the decoder to generate an output.
- retrieved documentproduces · A retrieved document is information selected by an information-retrieval process from a source and supplied as context or evidence for a query or task.
Questions about extractive summarization
What is extractive summarization?
Extractive summarization produces a shorter representation of source content by retaining selected original words, phrases, sentences, or passages and omitting other material.
Which Manning books explain extractive summarization?
Starting Data Analytics with Generative AI and Python, Retrieval Augmented Generation, The Foundational Ideas, Natural Language Processing in Action, Second Edition, Effective Conversational AI and Building LLM Applications with DSPy.
What should I know before learning extractive summarization?
Manning books that teach extractive summarization usually introduce language model, LLM-as-a-judge and embedding first.
What is the difference between extractive summarization and abstractive summarization?
Extractive summarization produces a shorter representation of source content by retaining selected original words, phrases, sentences, or passages and omitting other material. Abstractive summarization generates a concise account of source information using newly composed wording rather than only selecting text from the source.
What is extractive summarization a kind of?
Extractive summarization is a kind of summarization. Summarization is the production of a shorter representation of source information that preserves its important content or meaning.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.




