Microsoft Certified: Azure AI Engineer AssociateImplement natural language processing solutionsHard
A media company wants to implement a system that automatically generates short, engaging summaries of lengthy news articles and video transcripts for social media promotion. The summaries need to be concise, capture the main essence, and potentially rephrase content rather than just extracting sentences. Which Azure AI Language feature is best suited for this task?
- AExtractive Summarization
- BAbstractive Summarization
- CNamed Entity Recognition (NER)
- DKey Phrase Extraction
Show answer & explanationAnswer & explanation
Correct answer: B. Abstractive Summarization
Abstractive Summarization is designed to generate new, concise sentences that capture the core meaning of a document, often rephrasing the original content. This is ideal for 'engaging summaries' that 'rephrase content'. Extractive Summarization simply pulls existing sentences, which might not be as engaging or concise for social media. Key Phrase Extraction identifies important phrases, and NER extracts specific entities.
Why the other options are wrong
- A. Extractive Summarization extracts key sentences directly from the source, which may not be as concise or rephrased for engagement.
- C. NER identifies specific entities (e.g., names, locations), not a summary of the entire content.
- D. Key Phrase Extraction identifies important phrases, but does not generate a coherent summary.
Abstractive Summarization
An Azure AI Language feature that generates new, concise sentences to capture the main idea of a document, often rephrasing the original content.
- Creates novel sentences, not just extracts them.
- Produces more human-like and coherent summaries.
- Requires advanced natural language generation capabilities.
Memory trick: Abstractive is like a smart editor, creating new, punchy headlines, not just cutting paragraphs.