Courseiva

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

A customer service team wants to analyze thousands of support tickets to automatically categorize them into predefined topics like 'billing', 'technical issue', and 'account management'. They have a small set of labeled tickets for each category. Which Azure AI Language feature should they use?

⚠ Common exam trap

Many candidates confuse Custom Text Classification with pre-built features like Sentiment Analysis or Key Phrase Extraction, assuming any NLP feature can categorize tickets, but only Custom Text Classification allows training on your own labeled data for custom categories.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

Custom Text Classification

The Custom Text Classification feature of Azure AI Language allows you to train a model using a small set of labeled tickets to automatically categorize text into predefined topics like 'billing', 'technical issue', and 'account management'. This feature is specifically designed for scenarios where you have labeled data and need to classify documents into custom categories, making it ideal for analyzing support tickets.

Answer analysis

Option-by-option breakdown

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

  • Pre-built Text Classification

    Why it's wrong here

    Pre-built Text Classification refers to Azure's out-of-the-box language models, which are pre-trained on fixed taxonomies such as sentiment, language, and named entities; they cannot be re-trained or expanded to recognize a company's custom topics. These models perform generic detection but do not accept user-defined labels like 'billing issue' or 'hardware problem,' so they fail to meet the requirement for classifying tickets into business-specific categories. No pre-built Azure service offers a ready-made multi-topic classifier with your own categories, so this option is incorrect.

    When this WOULD be correct

    In a question asking about the general-purpose Azure AI service for understanding and analyzing text without custom training, such as 'Which service provides pre-built sentiment analysis and key phrase extraction?', option A might be the correct answer if it refers to Azure AI Language's pre-configured features.

  • Sentiment Analysis

    Why it's wrong here

    Sentiment Analysis scores text on an emotional spectrum from positive to negative to neutral, and can also flag mixed sentiment at the sentence and document level. It is specifically designed to gauge opinion and tone, not to identify the underlying subject matter, so a negative ticket about 'billing' and a negative ticket about 'technical issue' would both simply be 'negative.' Since the team needs topic-based classification rather than polarity scoring, this option does not fulfill the requirement.

    When this WOULD be correct

    This option would be correct if the question asked for extracting specific entities like product names or dates from support tickets, using a pre-built entity recognition model without custom training.

  • Custom Text Classification

    Why this is correct

    Custom Text Classification is a feature of Azure AI Language that lets you build a tailor-made model by training on a labeled set of support-ticket excerpts. After you upload examples tagged as 'billing' or 'technical issue,' the service learns distinctive word patterns and assigns your own category labels to new tickets. Because it handles multi-class or single-class output, it scales to thousands of incoming conversations and directly solves the team's need for topic-based analysis.

  • Key Phrase Extraction

    Why it's wrong here

    Key Phrase Extraction uses Azure AI Language's pre-built model to pull out the most relevant words and noun phrases — such as 'refund,' 'login error,' or 'slow performance' — from each ticket. It returns a list of keywords per document rather than assigning a single category, so it does not group tickets into 'billing' versus 'technical issue' buckets. While useful as a step toward analysis, it cannot by itself power categorical reporting or automated routing, making it the wrong tool for this scenario.

    When this WOULD be correct

    Option D would be correct if the question asked about extracting intents and entities from customer chat or voice conversations, where the goal is to understand user goals (e.g., 'reset password') and extract relevant details (e.g., account ID) from natural language interactions.

Option-by-option analysis

Why each answer is right or wrong

Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The AI-900 exam frequently reuses these exact scenarios with slightly different constraints.

Custom Text ClassificationCorrect answer

Why this is correct

Custom Text Classification is a feature of Azure AI Language that lets you build a tailor-made model by training on a labeled set of support-ticket excerpts. After you upload examples tagged as 'billing' or 'technical issue,' the service learns distinctive word patterns and assigns your own category labels to new tickets. Because it handles multi-class or single-class output, it scales to thousands of incoming conversations and directly solves the team's need for topic-based analysis.

Pre-built Text ClassificationWrong answer — click to see why

Why this is wrong here

Option A is not a valid Azure AI Language feature; the correct feature for custom text classification with predefined topics is Custom Text Classification (part of Azure AI Language).

★ When this WOULD be the correct answer

In a question asking about the general-purpose Azure AI service for understanding and analyzing text without custom training, such as 'Which service provides pre-built sentiment analysis and key phrase extraction?', option A might be the correct answer if it refers to Azure AI Language's pre-configured features.

Why candidates choose this

Candidates may confuse the generic name 'Azure AI Language' with the specific custom classification feature, or they might think any text analysis task uses the same service without considering the need for custom training.

Sentiment AnalysisWrong answer — click to see why

Why this is wrong here

This option is wrong because the question requires a custom text classification model trained on labeled data, not a pre-built entity extraction or language detection feature.

★ When this WOULD be the correct answer

This option would be correct if the question asked for extracting specific entities like product names or dates from support tickets, using a pre-built entity recognition model without custom training.

Why candidates choose this

Candidates may confuse entity extraction with text classification, thinking that extracting predefined topics is similar to extracting entities, or they may not understand the need for custom training.

Key Phrase ExtractionWrong answer — click to see why

Why this is wrong here

The question requires categorizing tickets into predefined topics using labeled data, which is a text classification task. Option D (e.g., 'Conversational Language Understanding') is designed for intent and entity extraction from conversational utterances, not for categorizing long-form text like support tickets into fixed categories.

★ When this WOULD be the correct answer

Option D would be correct if the question asked about extracting intents and entities from customer chat or voice conversations, where the goal is to understand user goals (e.g., 'reset password') and extract relevant details (e.g., account ID) from natural language interactions.

Why candidates choose this

Candidates may confuse categorization with intent recognition, as both involve classifying text. The mention of 'support tickets' and 'topics' might lead them to think of conversational AI features, but the key difference is that tickets are static documents, not interactive conversations.

Analysis generated from the official AI-900blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”

About these practice questions

Courseiva writes every AI-900 question from scratch — 985 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →

How Courseiva writes practice questions · Editorial policy

JA

Written by Johnson Ajibi, MSc IT Security

Senior Network & Security Engineer · founder of Courseiva

This AI-900 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI-900 exam.