Microsoft Practice Questions, Discussions & Exam Topics by our Authors
DRAG DROP
You plan to deploy an Azure Machine Leaming model by using the Machine Learning designer.
Which four actions should you perform in sequence?
-
You plan to deploy an Azure Machine Learning model by using the Machine Learning designer.
Which four actions should you perform in sequen...
Let’s carefully check the options and match them with the correct ML workflow:
The correct sequence for using Azure ML Designer is:
1. Ingest and prepare a dataset – bring in your data and clean/transform it.
2. Split the data randomly into training data and validation data – so you can train and test properly.
3. Train the model – using the training dataset.
4. Evaluate the model against the validation dataset – to see how well it performs on unseen data.
---
Now, check the options:
...
Author: NightmareDragon2025 · Last updated May 13, 2026
HOTSPOT
-
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
...
Let’s go carefully step by step, because these are subtle distinctions between clustering (unsupervised learning) and classification/prediction (supervised learning):
1. Organizing documents into groups based on different usage statistics
This is grouping items based on patterns in the data without predefined labels.
✅ This is clustering.
2. Grouping similar patients based on symptoms and diagnostic test results
...
Author: Daniel · Last updated May 13, 2026
HOTSPOT
When building a regression model, labels must have a data type of __________
-
Select the a...
Let’s carefully analyze the question:
> "When building a regression model, labels must have a data type of ___."
Key points:
Regression predicts a continuous numeric value (e.g., price, temperature, sales).
The label (the thing you are predicting) must therefore be numeric.
Regression cannot predict text, boolean, or datetime directly.
...
Author: Matthew · Last updated May 13, 2026
You need to create a clustering model and evaluate the model by using Azure Machine Learning designe...
The correct answer is B) Use the original dataset for training and evaluation. Here's the reasoning behind this choice and why the other options are rejected:
A) Split the original dataset into a dataset for training and a dataset for testing. Use the testing dataset for evaluation.
This approach is common in supervised learning models (such as classification and regression), where the model is trained on one part of the data (training dataset) and then evaluated on a separate part (testing dataset). However, clustering is an unsupervised learning method, and it does not rely on predefined labels or a separate test set for evaluation. The evaluation in clustering is often done based on how well the model groups the data rather than comparing predictions to labels.
Why rejected: Splitting the dataset into training and testing datasets is more appropriate for supervised learning tasks, not unsupervised clustering. Clustering doesn't have predefined labels to validate against.
Use case: This approach is used in supervised learning, not clustering, where labeled data is available for training and testing.
B) Use the original dataset for training and evaluation
In clustering, since it's an unsupervised learning method, the model can be trained and evaluated using the entire dataset. The goal of clustering is to find inherent patterns or groupings in the data without needing separate training and testing datasets. Evaluation is based on metrics such as silhouette score, Davies-Bouldin index, or internal cohesion rather than comparison with separate test data.
Why selected: For clustering, you generally don't need a separate testing dataset, as the goal is to group the entire dataset. The evaluation is typically done using internal metrics on the full dataset, making this the correct approach.
Use case: This is the standard approach in unsup...
Author: Lucas · Last updated May 13, 2026
You have a dataset that contains the columns shown in the following table.
You have a machine learning model that predicts the value of ColumnE b...
The correct answer is C) regression. Here's the reasoning behind this choice and why the other options are rejected:
A) Analysis
The term "analysis" is very broad and not specific to a type of machine learning model. While it could refer to any process of examining data, it does not define the machine learning technique used to predict values. In machine learning, "analysis" could refer to exploratory data analysis or statistical analysis, which are steps in the data preprocessing phase rather than the modeling phase itself.
Why rejected: "Analysis" does not refer to a specific machine learning model type. It is more of a general concept and does not directly involve predicting values, which is the main focus here.
Use case: "Analysis" could refer to initial data exploration, statistical analysis, or data visualization, but it isn't a machine learning model.
B) Clustering
Clustering is an unsupervised learning technique where the goal is to group similar data points together. It is typically used when you want to find patterns or groupings in data without predefined labels or values. Clustering algorithms do not predict a specific numeric value but rather group data into clusters based on similarity.
Why rejected: Clustering does not predict a specific target variable like Col...
Author: Arjun · Last updated May 13, 2026
You need to track multiple versions of a model that was trained by using Azure Machine Learning.
...
To track multiple versions of a model in Azure Machine Learning, the best approach is to register the model.
Explanation of the Selected Option:
B) Register the model is the correct option because registering the model allows you to keep track of different versions of the model. Each time a new model is trained, it can be registered with a version number. This makes it easy to manage multiple versions, compare them, and deploy the most appropriate version to production. The versioning mechanism is built into Azure Machine Learning when registering the model.
- Key factors:
- Version control: Each registered model can have a version number that is automatically incremented when new versions are registered.
- Tracking and management: Azure ML provides tools to help track models, including metadata like performance metrics, training data, and configuration settings. This is crucial for model management and for ensuring reproducibility.
- Scenario: Registering models is necessary when multiple versions of a model are being iterated on, as you can retrieve, compare, and deploy them easily.
Why the Other Options Are Rejected:
1. A) Explain the model: This option focuses on explaining or interpreting the model, which is a process ...
Author: Sofia · Last updated May 13, 2026
You need to identify groups of rows with similar numeric values in a dataset.
Which type of machi...
To identify groups of rows with similar numeric values in a dataset, the appropriate machine learning technique is clustering.
Explanation of the Selected Option:
A) Clustering is the correct option because clustering is an unsupervised machine learning technique used to group data points (rows) that are similar to each other based on certain features, such as numeric values. In clustering, the algorithm attempts to find natural groupings or patterns in the data without using predefined labels.
- Key factors:
- Grouping similar data: Clustering algorithms, such as K-means or hierarchical clustering, are designed to identify similarities between rows in the dataset and assign them to different groups or clusters based on their numeric features.
- Unsupervised learning: Clustering doesn't require labeled data and can be used when you do not have predefined labels for the data.
- Scenario: Clustering is particularly useful when you want to discover inherent groupings in the data, such as segmenting customers based on purchasing behavior or grouping sensor readings with similar patterns.
Why the Other Options Are Rejected:
1. B) Regression: Regression is a...
Author: Oscar · Last updated May 13, 2026
HOTSPOT
Select the answer that correctly completes the sentence.
A banking system that predicts whether a loan will be repaid is...
Let’s analyze this carefully.
Scenario:
Task: Predict whether a loan will be repaid.
Output: The answer is yes or no → a category.
---
Option analysis:
A. Clustering → Groups data without labels. ❌ Not suitable because we know the outcome (repaid or not).
B. Regression → Predicts numbers/continuous values...
Author: Grace · Last updated May 13, 2026
HOTSPOT
__________models can be used to predict the sale price of auctioned items
-
Select the answ...
The question is:
> “__________ models can be used to predict the sale price of auctioned items.”
We need to figure out what type of machine learning model fits.
---
Step 1: Identify the type of output
Predicting sale price → the output is a numerical value (e.g., $120, $500, etc.).
Numerical/continuous outputs are handled by regression models.
---
Step 2: Compare with other options
1. Classification
Used when the output is a category (e.g., “expensive” vs “cheap”).
❌ Not suitable because price is not a category, it’s a number.
2. Clusteri...
Author: NebulaEagle11 · Last updated May 13, 2026
A historian can use ________ to digitize newspaper articles.
Select the answer that correctly com...
To digitize newspaper articles, the correct answer is D) Optical character recognition (OCR).
Explanation of the Selected Option:
D) Optical character recognition (OCR) is the correct option because OCR technology is specifically designed to convert printed or handwritten text from scanned images (such as newspaper articles) into machine-readable text. OCR analyzes the structure of the document, identifies characters, and converts them into a digital format, allowing the historian to work with the content of the newspaper articles in a searchable and editable format.
- Key factors:
- Text recognition: OCR is ideal for recognizing printed text and converting it into digital formats, making it a perfect fit for newspaper articles, which are typically made up of text in various fonts and styles.
- Document digitization: OCR enables efficient digitization of historical documents, making it easier to store, search, and analyze large volumes of text from scanned images or PDFs.
- Scenario: Historians or archivists working with physical newspapers or documents can use OCR to convert old newspaper articles into digital files for easier access, preservation, and analysis.
Why the Other Options Are Rejected:
1. A) Object detection: Object detection is used to identify and ...
Author: Rahul · Last updated May 13, 2026
HOTSPOT
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
NOTE: Each correct selection is worth one point.
Object detection can identify the location of a damaged product in an image?.
Object detection can identify multiple inst...
---
Step-by-step explanation:
1. “Object detection can identify the location of a damaged product in an image?”
Object detection not only detects what the object is, but also where it is in the image by drawing bounding boxes around it.
✅ Answer: Yes
2. “Object detection can identify multiple instances of a damaged product in an image?”
Object detection can detect all occ...
Author: Vivaan · Last updated May 13, 2026
You need to create a model that labels a collection of your personal digital photographs.
Which Azu...
To label a collection of personal digital photographs, the correct service to use from Azure Cognitive Services is B) Custom Vision.
Explanation of the Selected Option:
B) Custom Vision is the correct option because it allows you to build custom image classification models tailored to your specific needs, such as labeling your personal photographs. With Custom Vision, you can upload your images, label them based on your preferences, and train a model that can automatically classify new images into these predefined labels.
- Key factors:
- Image classification: Custom Vision is specifically designed for training custom models to classify and label images based on categories you define. It is ideal for personalizing image labeling tasks, such as sorting photographs into categories like "vacation", "family", "landscape", etc.
- Customization: Unlike general pre-built services, Custom Vision allows you to create models that are highly specific to your dataset, so the model will be more accurate in labeling images that reflect your unique needs.
- Scenario: You can upload your personal collection of digital photos, label them according to your preferences, and then train the model to automatically classify and label new images.
Why the Other Options Are Rejected:
1. A) Form Recognizer: Form Recognizer is designed to extract information from structured documents...
Author: Daniel · Last updated May 13, 2026
HOTSPOT
__________is used to identify multiple types of items in one image.
-
Select the answer tha...
Let’s carefully reason this out.
The task: “Identify multiple types of items in one image.”
Key points:
Multiple items in the same image → need to detect locations of each item.
Different types → need to recognize different classes (e.g., widget, gadget, tool).
Now the options:
A. Object detection – ✅ Detects multiple objects, their types, and their locations in an ...
Author: ThunderBear · Last updated May 13, 2026
HOTSPOT
Identifying whether a kiosk user is annoyed by monitoring a video feed from the kiosk is an example of___________
-
S...
Let’s reason carefully.
The task: “Identifying whether a kiosk user is annoyed by monitoring a video feed from the kiosk.”
Key points:
You are analyzing emotions/expressions.
The goal is analyzing the face to infer a state, not just detecting or recognizing the person.
Now the options:
A. Face detection – ✅ Detects that a face exists in an image/video. ❌ Doesn’t a...
Author: Michael · Last updated May 13, 2026
DRAG DROP
-
Match the Azure Cognitive Services to the appropriate actions.
To answer, drag the appropriate service from the column on the left to its action on the right. Each service may be used ...
The question is asking to match tasks to the appropriate Azure Cognitive Service:
1. Identify objects in an image → This is about recognizing things like cars, chairs, or other items in an image.
Azure service: Custom Vision.
2. Automatically import data from an invoice to a database → This is about extracting structured information from forms or documents.
Azure service: Form Recognizer.
3. Identify people in an image → This is about detecting or verifying h...
Author: Ahmed97 · Last updated May 13, 2026
HOTSPOT
An Al solution that helps photographers take better portrait photographs by providing feedback on exposure, noise, and occlusion is an example of facial ...
The correct answer is: A) analysis ✅
Here’s why:
Facial detection → Simply identifies where a face is in an image. It doesn’t give detailed feedback.
Facial recognition → Identifies who the person is by comparing to a database.
Facial analysis → Extracts information a...
Author: Arjun · Last updated May 13, 2026
Your company manufactures widgets.
You have 1,000 digital photos of the widgets.
You need to identify the location of ...
To identify the location of the widgets within the photos, the correct service to use from Azure Cognitive Services is B) Custom Vision object detection.
Explanation of the Selected Option:
B) Custom Vision object detection is the correct option because it allows you to train a model that can detect and locate objects (in this case, widgets) within images. Object detection not only classifies objects but also provides the coordinates (bounding boxes) of where the objects are located in each image.
- Key factors:
- Object detection: Custom Vision's object detection feature is specifically designed to locate and identify objects within images, making it the ideal choice when you need both classification and localization (coordinates of the widgets within the photos).
- Customization: Since you are working with widgets, a custom-trained model will be more accurate, as it can be trained specifically on your widget images.
- Scenario: In this case, your task is to identify the exact position of the widgets within the photos, and object detection provides this capability along with classification (i.e., the model can detect the widget and also tell you where in the photo it appears).
Why the Other Options Are Rejected:
1. A) Computer Vision Spatial Analysis: Spatial analysis is generally used for analyzing the...
Author: FrostFalcon88 · Last updated May 13, 2026
You need to convert handwritten notes into digital text.
Which type of computer vision should you...
To convert handwritten notes into digital text, the correct option is B) Optical character recognition (OCR).
Explanation of the Selected Option:
B) Optical character recognition (OCR) is the correct choice because OCR is specifically designed to recognize text within images, including both printed and handwritten text. OCR technology analyzes the handwritten notes, identifies the characters, and converts them into machine-readable text.
- Key factors:
- Text recognition: OCR excels at recognizing and converting text from images into digital form, which is exactly the task required to convert handwritten notes into digital text.
- Handwriting support: OCR can also handle various handwriting styles, though its accuracy may vary depending on the legibility of the handwriting.
- Scenario: A person who has handwritten notes, such as meeting notes or lecture notes, can use OCR to convert them into digital text, making them easier to store, edit, and search.
Why the Other Options Are Rejected:
1. A) Facial detection: Facial detection focuses on identifying f...
Author: NebulaEagle11 · Last updated May 13, 2026
HOTSPOT
__________is used to identify multiple types of items in one image.
-
Select the answer tha...
Let’s reason carefully.
The task: “Identify multiple types of items in one image.”
Key points:
Multiple items in one image → Need locations and types for each object.
Now the options:
A. Image classification – Labels the whole image as a single category (e.g., “dog” or “cat”). ❌ Cannot detect multiple objects in the same image.
B. Image description...
Author: Oscar · Last updated May 13, 2026
You need to develop a mobile app for employees to scan and store their expenses while travelling.
W...
To develop a mobile app for employees to scan and store their expenses while traveling, the correct type of computer vision to use is D) Optical character recognition (OCR).
Explanation of the Selected Option:
D) Optical character recognition (OCR) is the most suitable option because OCR is specifically designed to extract text from images. In the case of a mobile app for scanning and storing expenses, OCR would allow the app to scan receipts or invoices, extract key text such as the total amount, date, merchant name, and other relevant details, and then store this information digitally.
- Key factors:
- Text extraction: OCR can recognize and extract printed or handwritten text from images, which is exactly what is needed for scanning receipts or invoices. The extracted text can then be used to automatically populate fields in the app, such as the amount and date of the expense.
- Expense tracking: OCR makes it possible to automate the data entry process, eliminating the need for manual input and reducing errors.
- Scenario: The app will need to capture and extract the relevant information from receipts or other expense-related documents. OCR is optimized for this task, as it can process documents, scan for text, and convert it into editable, machine-readable formats.
Why the Other Options Are Rejected:
1. A) Face detection: Face de...
Author: Daniel · Last updated May 13, 2026
HOTSPOT
You can use the __________service to train an object detection model by using your own images.
-
Select ...
The correct answer is:
B) Custom Vision ✅
Explanation:
Custom Vision is an Azure service that lets you train your own image classification or object detection models using your own labeled images. It’s ideal when you need a model tailored to your specific objects or scenarios.
Why the others are incorrect:
A) Computer ...
Author: Sam · Last updated May 13, 2026
HOTSPOT
An Al solution that helps photographers take better portrait photographs by providing feedback on exposure, noise, and occlusion is an example of facial __________
-
...
Let’s carefully analyze this:
The scenario:
AI helps photographers by analyzing portraits for exposure, noise, and occlusion.
It’s not identifying who the person is.
It’s not just finding a face in an image.
It’s analyzing the face for quality/features.
---
Definitions:
1. Facial Detection → Finding where faces are in an image.
Example: “There’s a face at coordinates...
Author: ThunderBear · Last updated May 13, 2026
DRAG DROP
-
Match the Azure Cognitive Services to the appropriate AI workloads.
To answer, drag the appropriate service from the column on the left to its workload on the right. Each service may be ...
Correct Answer: A) Custom Vision, Form Recognized, Face
Explanation
1. Identify objects in an image → Custom Vision
Custom Vision is used to build image classification and object detection models.
It allows you to train a model to recognize specific objects in images (for example: cars, products, defects).
2. Automatically import data from an invoice to a database → Form Recognizer
Form Recognizer (now called Azure AI Document Intelligence) extracts structured data from documents such as invoices, receipts, and forms.
It can automatically...
Author: Ella · Last updated May 13, 2026
You need to implement a pre-built solution that will identify well-known brands in digital photographs.
Which...
To implement a pre-built solution that will identify well-known brands in digital photographs, the correct service to use is D) Computer Vision.
Explanation of the Selected Option:
D) Computer Vision is the most appropriate option because it includes pre-built models capable of analyzing images and identifying various objects, including logos of well-known brands. The Computer Vision API can automatically detect and recognize objects, logos, and brands in images, making it ideal for tasks like identifying branded products or company logos in photographs.
- Key factors:
- Pre-built models: The Computer Vision API comes with pre-trained models that are capable of recognizing well-known logos and objects, meaning you do not need to train the model yourself.
- Object recognition: It is particularly good at analyzing the contents of images, including logos and brands, making it a perfect fit for identifying well-known brands.
- Scenario: In a scenario where you want to automatically identify logos in images (e.g., identifying brands in photos taken by consumers or employees), Computer Vision can do this efficiently without requiring any customization.
Why the Other Options Are Rejected:
1. A) Custom Visio...
Author: NightmareDragon2025 · Last updated May 13, 2026
Natural language processing can be used to __________.
Select the answer that correctly completes...
The correct completion for the sentence is C) Classify email messages as work-related or personal.
Explanation of the Selected Option:
C) Classify email messages as work-related or personal is the correct option because natural language processing (NLP) is specifically designed to process and analyze text data. NLP techniques can be used to analyze the content of email messages and classify them based on context, keywords, or patterns found in the text. By understanding the text within emails, an NLP model can distinguish between work-related and personal messages, making it ideal for tasks like email classification.
- Key factors:
- Text analysis: NLP focuses on understanding, processing, and interpreting human language in text form. It can be used for tasks like sentiment analysis, categorization, and text classification.
- Scenario: In an email filtering system, NLP can be used to automatically categorize emails based on their content and classify them into categories like "work-related" or "personal."
Why the Other Options Are Re...
Author: Sam · Last updated May 13, 2026
You plan to develop a bot that will enable users to query a knowledge base by using natural language processing.
Which two services should you include in the solution? Each correct answer ...
To develop a bot that enables users to query a knowledge base using natural language processing (NLP), the two most appropriate services would be:
A) Language Service
- Reasoning: The Azure Language Service offers NLP capabilities, such as language understanding (LUIS), text analytics, and translation. This service is directly relevant to processing and understanding the natural language input from users, allowing the bot to interpret user queries, extract meaning, and engage in a meaningful conversation. The Language Service is designed specifically for tasks like language comprehension, entity extraction, sentiment analysis, and question-answering, making it a crucial component in building an NLP-based bot.
B) Azure Bot Service
- Reasoning: The Azure Bot Service is a comprehensive platform to develop and deploy conversational bots. It enables you to create bots that can interact with users using various channels (e.g., web, mobile, Microsoft Teams). It integrates with other A...
Author: VioletCheetah55 · Last updated May 13, 2026
HOTSPOT
-
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
...
Let’s evaluate each statement again and match them to the options.
---
Statement 1
```
/translate?from-it&to=fr&to=en
```
`from=it` → input is Italian, not English
❌ No
---
Statement 2
```
/translate?from-en&to-fr&to-it
``...
Author: Isabella · Last updated May 13, 2026
An app that analyzes social media posts to identify their tone is an example of which type of natura...
The correct answer is A) Sentiment Analysis.
Reasoning:
- A) Sentiment Analysis: Sentiment analysis is a common NLP workload that focuses on identifying the sentiment or tone of a given text, such as social media posts. It analyzes whether the sentiment expressed in the text is positive, negative, or neutral, which is exactly the goal of analyzing the tone in social media posts. This process is highly relevant for determining how people feel about certain topics, brands, or products based on their online posts.
Why the other options are rejected:
- B) Speech Recognition: Speech recognition involves converting spoken language into written text. It is not related to analyzing the tone of text or social media posts. Speech recognition is more suited for applications where the input is spoken language (...
Author: Max · Last updated May 13, 2026
You are building a chatbot that will use natural language processing (NLP) to perform the following actions based on the text input of a user.
Accept customer orders.
Retrieve suppo...
For Microsoft Azure AI-900 (Azure AI Fundamentals), the correct answer is:
✅ D. language modeling
---
AI-900 exam reasoning (important 🎯)
In AI-900, language modeling is defined as an NLP workload that enables systems to:
Understand user intent
Generate appropriate responses
Support chatbots and conversational AI
Your scenario involves:
Accepting customer orders
Retrieving support documents
Retrieving order status...
Author: Rohan · Last updated May 13, 2026
DRAG DROP
-
Match the Azure Cognitive Services service to the appropriate actions.
To answer, drag the appropriate service from the column on the left to its action on the right. Each service may be use...
Let’s carefully match each task to the correct Azure service step by step:
---
Tasks:
1. Convert spoken requests into text
This is speech-to-text → Speech
2. Identify the intent of a user's requests
Figuring out what the user wants → Language Service
3. Apply intent to entities and utterances
Extract entities from text and map to intent → Language Service
---
Mapping to options:
| Option | Mapping ...
Author: Suresh · Last updated May 13, 2026
HOTSPOT
-
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
NOTE: Each correct selection is worth one point.
A webchat bot can interact with users visiting a website?
Automatically generating captions for pre-recorded videos is an example of natural language processing?.
A smar...
The correct AI-900 answer is:
✅ C) Yes, No, Yes
Explanation (short & exam-focused):
1. A webchat bot can interact with users visiting a website → Yes
Chatbots use Natural Language Processing (NLP) to understand and reply to user text.
2. Automatically generating captions for pre-record...
Author: Zara · Last updated May 13, 2026
You have a website that includes customer reviews.
You need to store the reviews in English and present the reviews to users in their respective language by recognizing each user's geographi...
The correct answer is D) Translation.
Reasoning:
- D) Translation: Since the requirement is to store the reviews in English and then present them to users in their respective languages based on geographical location, translation is the most appropriate NLP workload. Translation can convert the reviews from English into the languages required for different user groups, ensuring that the content is accessible in the user’s preferred language. This is a common use case for translation services, especially when dealing with multilingual websites or applications.
Why the other options are rejected:
- A) Key Phrase Extraction: Key phrase extraction involves identifying the most important words or phrases within a text to summarize its content. While this is useful for analyzing or categoriz...
Author: Leo · Last updated May 13, 2026
HOTSPOT
-
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
...
Let's carefully analyze each statement one by one in the context of AI-900: Microsoft Azure AI Fundamentals, specifically about Azure Bot Services and Chatbots:
1. Chatbots can support voice input?
Yes. Azure Bot Service can integrate with Speech Services, allowing users to interact via voice. ✅
2. A separate chatbot is required for each communication channel?
No. Azure Bot Service al...
Author: Chloe · Last updated May 13, 2026
HOTSPOT
-
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
NOTE: Each correct selection is worth one point.
A bot that responds to queries by internal users is an example of a natural language processing workload?.
A mobile application that displays images relating to an entered search term is an example of ...
Let's carefully analyze each scenario in the context of Natural Language Processing (NLP), which deals with understanding, interpreting, and generating human language.
1. A bot that responds to queries by internal users
This involves understanding user questions (text) and generating appropriate responses.
✅ This is an NLP workload.
2. A mobile application that displays images relating to an entered search term
The core functionality here is image retrieval b...
Author: John · Last updated May 13, 2026
You have a solution that analyzes social media posts to extract the mentions of city names and the city names discussed most frequently.
Which type of...
The correct answer is D) Entity Recognition.
Reasoning:
- D) Entity Recognition: Entity recognition (or named entity recognition - NER) is the process of identifying specific entities, such as names of people, places, organizations, and other key elements, in a text. In this scenario, the solution focuses on extracting mentions of city names, which are a type of entity. Entity recognition is well-suited for this task because it enables the system to identify and classify city names from the social media posts, making it an essential NLP workload for extracting the relevant geographical entities.
Why the other options are rejected:
- A) Speech Recognition: Speech recognition converts spoken language into written text. It is not relevant to the scenario at hand, as the focus is on extracting information from written text (social media posts), not...
Author: Sofia · Last updated May 13, 2026
HOTSPOT
-
For each of the following statements, select Yes if the statement is true. Otherwise, select No.
NOTE: Each correct selection is worth one point.
You can use Language Service's question answering to query an Azure SQL database?
You should use Language Service's question answering when you want a knowledge base to provide the ...
Let’s carefully analyze each statement:
1. “You can use Language Service's question answering to query an Azure SQL database”
Azure Language Service's question answering is designed to provide answers from a knowledge base or documents, not directly query databases like Azure SQL.
This is No.
2. “You should use Language Service's question answering when you want a knowledge base to provide the same answer to different users who submit similar question...
Author: Abigail · Last updated May 13, 2026
You are developing a solution that uses the Language service.
You need to identify the main talking points in a collection of documents.
...
The correct answer is D) Key Phrase Extraction.
Reasoning:
- D) Key Phrase Extraction: Key phrase extraction is an NLP technique used to identify the most important terms or phrases in a document that summarize its main talking points or themes. This is ideal for the scenario where you need to extract the core ideas or topics from a collection of documents. By identifying key phrases, the solution can highlight the central themes or subjects discussed across the documents, which aligns with the requirement to identify the main talking points.
Why the other options are rejected:
- A) Language Detection: Language detection identifies the language in which a piece of text is written. This is important for determining which language processing model to apply, but it does not help in extracting main talking points...
Author: Maya · Last updated May 13, 2026
DRAG DROP
-
You are designing a system that will generate insurance quotes automatically.
Match the Microsoft responsible AI principles to the appropriate requirements.
To answer, drag the appropriate principle from the column on the left to its requirement on the right. Each principle may be used once, more than once, or not at all.
NOTE: Each correct match is worth one point.
A customer's personal information must be visible only to staff who are involved in the decision-making proces...
Let's carefully break this down step by step:
You gave three requirements:
1. Customer’s personal information must be visible only to staff involved in decision-making
This is about protecting sensitive data, so it falls under Privacy and security. ✅
2. Decision-making process must be recorded so staff can identify reasoning behind a quote
This is about being able to see how decisions were made, which is Transparency. ✅
3. System must be accessible to customers who use screen readers or other assistive technolo...
Author: Aarav · Last updated May 13, 2026
Which type of natural language processing (NLP) entity is used to identify a phone number?
The correct answer is A) Regular Expression.
Reasoning:
- A) Regular Expression: Regular expressions (regex) are a powerful tool used to match patterns in text, making them ideal for identifying structured data like phone numbers. Phone numbers follow specific formats (e.g., (123) 456-7890 or +1-800-555-1234), which can be described using a regex pattern. By defining a pattern for phone numbers, regular expressions can accurately identify and extract them from a text. This is the most efficient and straightforward approach for identifying phone numbers.
Why the other options are rejected:
- B) Machine-learned: Machine-learned approaches are based on training models to identify entities from data. While machine learning is useful for complex, context-dependent entity recognition (e.g., identifying named en...
Author: Matthew · Last updated May 13, 2026
HOTSPOT
Returning a bounding box that indicates the location of a vehicle in an image is an example of __________
-
To complete the s...
Let’s analyze this carefully:
Scenario:
You have an image.
You want to return a bounding box that shows the location of a vehicle.
Now, check each option:
A) Image classification ❌
Image classification assigns a label to the entire image (e.g., “car” or “dog”).
It does not tell you where the object is.
B) Object detection...
Author: Daniel · Last updated May 13, 2026
Your company is exploring the use of voice recognition technologies in its smart home devices. The company wants to identify any barriers that might unintentionally leave out specific user gro...
In this scenario, the guiding principle being referred to is D) Inclusiveness.
Explanation:
The focus of this situation is on ensuring that the use of voice recognition technology in smart home devices does not unintentionally leave out certain user groups. This involves considering how different individuals, with varying abilities or characteristics, might interact with these devices and ensuring that all users, regardless of their specific needs or limitations, can access and benefit from the technology. This is at the heart of inclusiveness, which prioritizes designing AI systems in a way that they work for everyone, including those from underrepresented or vulnerable groups.
Why Other Options Are Rejected:
- A) Accountability: Accountability refers to the responsibility of developers and organizations in ensuring that their AI systems are transparent, answerable, and can be audited for their actions and decisions. While accountability is important, it does not directly address the inclusion of specific user groups in the context of voice recognition systems.
- B) Fairness: Fairness is about ensuring that AI systems do not exhibit bias and provide equal ...
Author: Elizabeth · Last updated May 13, 2026
HOTSPOT
-
You have a large dataset that contains motor vehicle sales data.
You need to train an automated machine learning (automated ML) model to predict vehicle sale values based on the type of vehicle.
Which task should you select? To ...
Let’s carefully analyze:
Goal: Predict vehicle sale values → this is a continuous numeric value, not a category.
Input: Vehicle type (categorical feature), maybe other features.
Match with the task types:
A) Classification → Predicts categories (e.g., red, blue, yes/no). ❌
B) Regression → Predicts continuous numeric va...
Author: SolarFalcon11 · Last updated May 13, 2026
HOTSPOT
Select the answer that correctly completes the sentence.
When evaluating the performance of a model, ___________the displays the predicted and ...
Let’s carefully go through this:
The description:
> “Displays the predicted and actual positives and negatives by using a grid of 0 and 1 values.”
A. AUC metric → Measures the area under the ROC curve, summarizing model performance. ❌
B. Confusion matrix → Shows true positives, true ...
Author: Alexander · Last updated May 13, 2026
You need to convert receipts into transactions in a spreadsheet. The spreadsheet must include the date of the transaction, the merchant, the total spe...
In this scenario, the most suitable Azure AI service would be B) Form Recognizer.
Explanation:
Form Recognizer is an AI-powered tool designed to extract data from forms, invoices, receipts, and other structured documents. It can automatically analyze the data from receipts, identify key fields such as date, merchant, total spent, and taxes paid, and extract them into a structured format like a spreadsheet. This makes it the most suitable choice for converting receipts into transactions with specific fields in a spreadsheet.
Why Other Options Are Rejected:
- A) Custom Vision: Custom Vision is primarily used for image classification and object detection tasks. It can be trained to recognize custom images, such as identifying specific objects or patterns within images. However, it is not specialized in processing forms or documents and would not be appropriate for extracting structured data like receipts.
- C) Face: The Face service is used for facial re...
Author: FrostFalcon88 · Last updated May 13, 2026
HOTSPOT
Select the answer that correctly completes the sentence.
Predicting how many vehicles will travel across a...
Let’s carefully analyze the scenario:
The task: predicting how many vehicles will travel across a bridge.
The output: a numerical value (number of vehicles).
Now match with machine learning types:
A. Classification → Predict...
Author: Abigail · Last updated May 13, 2026
HOTSPOT
-
Select the answer that correctly completes the sentence.
In a machine learning model, the data...
Let’s carefully break this down:
In machine learning:
Features → The input variables that the model uses to make predictions. ✅
Functions → Not used in this context. ❌
Labels → The ou...
Author: Abigail · Last updated May 13, 2026
You have a security system that analyzes images from CCTV to provide authorized staff entry into restricted area.
...
In this scenario, the most suitable type of computer vision would be C) Facial detection and facial recognition.
Explanation:
The system described analyzes images from CCTV to provide authorized staff entry into restricted areas. This directly suggests that the system is using facial recognition technology, which is specifically designed to identify and verify individuals based on their facial features. Facial recognition involves both facial detection (detecting the presence of a face) and facial recognition (matching the face to a stored database of authorized personnel). This is a common and effective use case for security systems where access control is based on identity verification.
Why Other Options Are Rejected:
- A) Optical Character Recognition (OCR): OCR is used for extracting text from images, such as reading printed text from documents or signs. While OCR could be useful in other contexts (e.g., reading labels or signs in a security system), it is not relevant for identifying people through CCTV images, which involves analyzing faces rather than text.
- B) Semantic Segmentation: Semantic segmentation is a type of com...
Author: Aarav · Last updated May 13, 2026
For which two workloads can you use computer vision? Each correct answer presents a complete solution.
N...
The two correct workloads for which you can use computer vision are:
A) Assigning the color pixels in an image to object names
- Explanation: This refers to semantic segmentation or object detection. These are computer vision techniques where each pixel in an image is classified into categories (e.g., assigning pixels of a car to the label "car"). Object detection and segmentation involve identifying and labeling different regions or objects in an image. This is a key application of computer vision to understand and categorize visual data.
E) Describing the contents of an image
- Explanation: This is related to image captioning and image classification, where computer vision models analyze an image and generate a description of what is present in the image. For example, recognizing that an image contains a "dog on a grassy field" or "a man riding a bicycle." It involves extracting high-level semantic understanding from an image, which is a fundamental use case of computer vision.
Why Other Options Are Rejected:
- B) Detecting inconsistencies and anomalies in a stream of data: This task is more related to anomaly detection in data streams or time-series data analysis. While computer vision may be involved in analyzing images or video feeds in specific scenarios, detecting inconsistencies in data streams is not a direct application of computer vision...
Author: Matthew · Last updated May 13, 2026
You have an app that identifies the coordinates of a product in an image of a supermarket shelf.
...
The most suitable service for this scenario is B) Custom Vision object detection.
Explanation:
The app described is identifying the coordinates of a product on a supermarket shelf. This suggests that the task involves object detection, which is a technique that allows an AI model to locate and identify multiple objects within an image, and return their coordinates (bounding boxes). Custom Vision object detection is the ideal service in this case because it specializes in training models to detect specific objects (such as products) within images, and it can output both the label of the object and the coordinates of where that object appears in the image.
Why Other Options Are Rejected:
- A) Custom Vision classification: This service is used for classifying images into predefined categories. However, it does not provide coordinates or bounding boxes for where the objects are located in the image. It simply categorizes the entire image, which is not sufficient for the need to identify the location of products on a shelf.
- C) Computer Vision Read...
Author: Liam · Last updated May 13, 2026
HOTSPOT
Select the answer that correctly completes the sentence.
A traffic monitoring system that collects vehicle registration numbers from CCTV fo...
Let’s carefully analyze this:
The system: collects vehicle registration numbers from CCTV footage.
Key point: it’s reading text (license plates) from images, not just detecting cars.
Now match with Computer Vision tasks:
A. Image classification → Labels the whole image (e.g., “car” or “truck”...
