AI-103 Practice Question 59: Key Phrase vs Named Entity
Microsoft Azure AI Apps and Agents Developer Associate (AI-103) · Test 6 · question 9 of 10
A media-monitoring service scans thousands of news articles a day and wants a short list of the main topics each article covers, for a headline-browsing screen: for example, an article about a factory closure should surface something like "factory closure" and "job losses," not a list of the people, companies, or dates the article happens to mention.
Which Azure Language feature fits this requirement?
Show answer and explanation
Answer:
Key phrase extraction is defined for exactly this job: "identify the main concepts in text." In Microsoft's own example, the sentence "The food was delicious and the staff were wonderful" returns "food" and "wonderful staff" (thematic phrases, not entities) which is the same shape of output the media-monitoring screen needs, and it works out of the box with no labeled training data.
- A is wrong because NER extracts specific typed entities ("person, location, event, product, and organization") not the general subject of the text. For the factory-closure example, NER would surface the company name and a date, not the concepts "factory closure" or "job losses." - C is wrong because PII detection identifies and redacts sensitive personal data; it isn't designed to summarize what an article is about, and "configuring custom categories" for topics repurposes a redaction feature for a job it doesn't do. - D is wrong because custom classification requires a labeled training set the team would have to build and maintain, when the preconfigured key phrase feature needs no customization or training data at all.