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.

• ✗

  A tool for counting the number of words and characters in a document

  Why it's wrong here

  A word/character counter simply measures document length by counting tokens and symbols—a deterministic, non-AI operation. It does not interpret the document's meaning or context. Azure AI Language text analytics, in contrast, uses natural language processing models to classify sentiment, extract key phrases and entities, and identify language, converting unstructured text into actionable insights.

• ✓

  AI-powered extraction of insights (sentiment, key phrases, entities) from unstructured text

  Why this is correct

  Correct—text analytics in Azure AI Language automates the extraction of insights from unstructured text using pre-trained natural language processing models. It returns sentiment scores (positive, negative, mixed, or neutral), identifies key phrases, recognizes named entities such as people, organizations, and locations, and can detect the document's language. This is precisely what Microsoft's text analytics (Azure AI Language) is designed to do.

• ✗

  A spreadsheet function for analysing numerical data in text cells

  Why it's wrong here

  Spreadsheet functions operate on structured cells with deterministic formulas—for example, SUM or AVERAGE—and lack the ability to understand human language context or semantics. Even when a cell contains text, a spreadsheet formula can only pattern-match or reformat it, not identify whether the language is positive/negative or what entities are mentioned. Text analytics is an Azure AI service that applies machine learning to unstructured text, which is a fundamentally different and much broader capability than numerical spreadsheet calculations.

• ✗

  Encrypting sensitive text data before storing it in the cloud

  Why it's wrong here

  Encryption is a security control that obfuscates text using algorithms and keys so unauthorized parties cannot read it; it does not analyze, interpret, or extract meaning from content. Text analytics requires access to readable plain text to apply NLP models—if the text is encrypted before being stored, the service cannot perform sentiment analysis, entity extraction, or key phrase detection. Encryption protects data; text analytics derives insight.

Go deeper

Related to this question