AI-900 Practice Question: Describe features of Natural Language Processing workloads on Azure

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  Extractive writes shorter summaries; abstractive writes longer ones

  Why it's wrong here

  This statement confuses summary length with method. Both extractive and abstractive summaries in Azure AI Language can be configured to produce shorter or longer outputs by adjusting parameters such as maximum summary length or sentence count. The fundamental distinction is whether the summary consists of copied sentences from the original or newly generated ones, not the relative length. Therefore, length is a tunable setting, not a defining characteristic of either summarization type.

• ✓

  Extractive selects key sentences verbatim; abstractive generates new sentences capturing the meaning

  Why this is correct

  This correctly describes the core difference: extractive summarization selects existing sentences directly from the input text and outputs them verbatim, based on a scoring model that ranks sentence importance. Abstractive summarization uses generative AI to produce new sentences, often paraphrased, that capture the document's key meaning while not necessarily appearing in the original. This is why abstractive summarization requires deeper natural-language understanding and generation capabilities, making this the accurate answer.

• ✗

  Abstractive summarisation is only available for non-English languages

  Why it's wrong here

  This is false because Azure AI Language provides both extractive and abstractive summarization for multiple languages, not exclusively for non-English text. In fact, English is typically the most fully supported language, with additional languages available depending on the specific feature and service version. The availability of abstractive summarization is determined by the language's model support, not by a restriction to non-English languages, so this statement misrepresents the service's actual capabilities.

• ✗

  Extractive summarisation uses generative AI; abstractive uses keyword ranking

  Why it's wrong here

  This statement reverses the actual mechanics. Extractive summarization in Azure AI Language scores and ranks sentences from the original document, then copies the highest-ranking sentences verbatim into the summary—no new text is generated. Abstractive summarization, by contrast, uses generative neural models to create entirely new sentences that paraphrase and condense the source content. Therefore, extractive relies on selection, not generative AI, while abstractive is the generative method.

Go deeper

Related to this question