Microsoft AI-103 Practice Test 79 - Prepaway
Practice Exams:

Pass Microsoft AI-103 Exam in First Attempt Guaranteed!

Get 100% Latest Exam Questions, Accurate & Verified Answers!

30 Days Free Updates, Instant Download!

Verified By Experts
Premium File
$89.99 $98.99

AI-103 Premium ETE File


  • Premium File 187 Questions & Answers. Last Update: Nov 15, 2026
Whats Included:
  • Latest Questions
  • 100% Accurate Answers
  • Fast Exam Updates

Microsoft AI-103 Practice Test 79

Question 780

Situation: A city licensing office receives abbreviations that map to different local agencies depending on context. Which TWO proposed practices create material safety, correctness, or reliability risks? Select TWO answers.

  1. Inspect records that explain the affected process, then check the result described in this case: A city licensing office receives abbreviations that map to different local agencies depending on context
  2. Define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review
  3. Treat every capitalized word as an organization entity
  4. Invent missing account numbers to satisfy a required field
  5. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented

Explanation

Correct Answer

C:

Because a city licensing office receives abbreviations that map to different local agencies depending on context, “Treat every capitalized word as an organization entity” would violate the requirement. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

D:

“Invent missing account numbers to satisfy a required field” would violate the requirement. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

Incorrect Answers

A:

“Inspect records that explain the affected process, then check the result described in this case: A city licensing office receives abbreviations that map to different local agencies depending on context” adds a relevant safeguard or verification step rather than weakening the solution. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

B:

“Define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review” is the controlled implementation, not an unsafe workaround. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Confirm that extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

E:

Because a city licensing office receives abbreviations that map to different local agencies depending on context, “Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented” is the required outcome of a working control, not an action that weakens it. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Confirm that extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

Question 781

Architecture case: A media archive identifies product and organization mentions in articles, but names can overlap with common words. What configuration should the engineer choose for this workload?

  1. Discard the source span that supports an entity label
  2. Normalize multiple unrelated people to one identifier
  3. Define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review
  4. Apply the proposed setting to all connected workloads before testing the case-specific boundary: a media archive identifies product and organization mentions in articles, but names can overlap with common words

Explanation

Correct Answer

C:

Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Confirm that extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

Incorrect Answers

A:

“Discard the source span that supports an entity label” does not implement the control that the facts call for. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Instead, define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review.

B:

Because a media archive identifies product and organization mentions in articles, but names can overlap with common words, “Normalize multiple unrelated people to one identifier” does not implement the control that the facts call for. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Instead, define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review.

D:

“Apply the proposed setting to all connected workloads before testing the case-specific boundary: a media archive identifies product and organization mentions in articles” does not implement the control that the facts call for. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Instead, define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review.

Question 782

Production requirement: A media archive identifies product and organization mentions in articles, but names can overlap with common words. What technical fact prevents the simplest shortcut from being reliable?

  1. Successful provisioning by itself proves all security, output, and operational constraints are satisfied
  2. The fact that the team started work is enough to prove correctness for this exact case: a media archive identifies product and organization mentions in articles, but names can overlap with common words
  3. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain
  4. The most powerful or least restricted model is always the safest choice regardless of task conditions

Explanation

Correct Answer

C:

Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review.

Incorrect Answers

A:

“Successful provisioning by itself proves all security, output, and operational constraints are satisfied” does not fully explain the decisive technical boundary in this case. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

B:

“The fact that the team started work is enough to prove correctness for this exact case: a media archive identifies product and organization mentions in articles, but names can overlap with common words” does not fully explain the decisive technical boundary in this case. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. The implementation must define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review.

D:

A media archive identifies product and organization mentions in articles, but names can overlap with common words. “The most powerful or least restricted model is always the safest choice regardless of task conditions” does not fully explain the decisive technical boundary in this case. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. The implementation must define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review.

Question 783

A team is investigating this task: A media archive identifies product and organization mentions in articles, but names can overlap with common words. What end-to-end test would distinguish successful operation from an installed configuration?

  1. The operator reports success without reproducing the observed behavior: a media archive identifies product and organization mentions in articles, but names can overlap with common words
  2. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented
  3. A project dashboard remains green, although the affected request, output and permissions were not checked
  4. The team closes a change ticket without replaying a representative positive and negative test

Explanation

Correct Answer

B:

A media archive identifies product and organization mentions in articles, but names can overlap with common words. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

Incorrect Answers

A:

“The operator reports success without reproducing the observed behavior: a media archive identifies product and organization mentions in articles, but names can overlap with common words” alone does not establish all the required behavior. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

C:

A media archive identifies product and organization mentions in articles, but names can overlap with common words. “A project dashboard remains green, although the affected request, output and permissions were not checked” would not meet the requirement. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

D:

Because a media archive identifies product and organization mentions in articles, but names can overlap with common words, “The team closes a change ticket without replaying a representative positive and negative test” alone does not establish all the required behavior. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

Question 784

Situation: A media archive identifies product and organization mentions in articles, but names can overlap with common words. Which diagnostic investigation best isolates the issue before another change?

  1. Read only the broad service-health summary and skip actual workload evidence
  2. Assume the general service is at fault without checking the affected operation in this case: a media archive identifies product and organization mentions in articles, but names can overlap with common words
  3. Compare source offsets, label definitions, confidence or evidence, schema errors and human adjudication examples
  4. Restart every affected component before collecting request or tool traces

Explanation

Correct Answer

C:

Because a media archive identifies product and organization mentions in articles, but names can overlap with common words, compare source offsets, label definitions, confidence or evidence, schema errors and human adjudication examples. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

Incorrect Answers

A:

A media archive identifies product and organization mentions in articles, but names can overlap with common words. “Read only the broad service-health summary and skip actual workload evidence” would not meet the requirement. Compare source offsets, label definitions, confidence or evidence, schema errors and human adjudication examples. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

B:

“Assume the general service is at fault without checking the affected operation in this case: a media archive identifies product and organization mentions in articles” would not meet the requirement. Compare source offsets, label definitions, confidence or evidence, schema errors and human adjudication examples. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

D:

A media archive identifies product and organization mentions in articles, but names can overlap with common words. “Restart every affected component before collecting request or tool traces” would not be the most direct first check for this failure. Compare source offsets, label definitions, confidence or evidence, schema errors and human adjudication examples.

Question 785

Architecture case: A media archive identifies product and organization mentions in articles, but names can overlap with common words. Which TWO shortcuts should be rejected for this scenario? Select TWO answers.

  1. Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented
  2. Invent missing account numbers to satisfy a required field
  3. Treat every capitalized word as an organization entity
  4. Define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review
  5. Inspect records that explain the affected process, then check the result described in this case: A media archive identifies product and organization mentions in articles, but names can overlap with common words

Explanation

Correct Answer

B:

Because a media archive identifies product and organization mentions in articles, but names can overlap with common words, “Invent missing account numbers to satisfy a required field” would violate the requirement. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

C:

A media archive identifies product and organization mentions in articles, but names can overlap with common words. “Treat every capitalized word as an organization entity” would violate the requirement. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

Incorrect Answers

A:

“Extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented” is the required outcome of a working control, not an action that weakens it. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

D:

Because a media archive identifies product and organization mentions in articles, but names can overlap with common words, “Define the allowed entity schema and normalization rules, validate extracted spans against source text and route ambiguous cases for review” is the controlled implementation, not an unsafe workaround. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain.

E:

“Inspect records that explain the affected process, then check the result described in this case: A media archive identifies product and organization mentions in articles” adds a relevant safeguard or verification step rather than weakening the solution. Entity labels are not reliable unless type definitions, span evidence and normalization behavior are consistent across the domain. Confirm that extracted entities match source spans, required categories use approved canonical values and ambiguous mentions are flagged rather than invented.

Question 786

Production requirement: A hotel guest praises the room but strongly criticizes check-in time in the same survey response. Which next change most directly addresses the reported technical risk?

  1. Ignore negation and sarcasm in quality review
  2. Collapse opposite aspects into one global sentiment label
  3. Apply the proposed setting to all connected workloads before testing the case-specific boundary: a hotel guest praises the room but strongly criticizes check-in time in the same survey response
  4. Evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions

Explanation

Correct Answer

D:

Evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions. Confirm that the model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages. If the result fails, whole-document sentiment can hide opposite opinions about different products, and literal polarity is unreliable for sarcasm.

Incorrect Answers

A:

“Ignore negation and sarcasm in quality review” would not meet the requirement. Instead, evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions.

B:

“Collapse opposite aspects into one global sentiment label” does not implement the control that the facts call for. Instead, evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions.

C:

“Apply the proposed setting to all connected workloads before testing the case-specific boundary: a hotel guest praises the room but strongly criticizes check-in time in the same survey response” does not implement the control that the facts call for. Instead, evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions.

Question 787

A team is investigating this task: A hotel guest praises the room but strongly criticizes check-in time in the same survey response. Which explanation best captures the key behavior of this architecture?

  1. Whole-document sentiment can hide opposite opinions about different products, and literal polarity is unreliable for sarcasm
  2. The fact that the team started work is enough to prove correctness for this exact case: a hotel guest praises the room but strongly criticizes check-in time in the same survey response
  3. The most powerful or least restricted model is always the safest choice regardless of task conditions
  4. Successful provisioning by itself proves all security, output, and operational constraints are satisfied

Explanation

Correct Answer

A:

Whole-document sentiment can hide opposite opinions about different products, and literal polarity is unreliable for sarcasm. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

Incorrect Answers

B:

“The fact that the team started work is enough to prove correctness for this exact case: a hotel guest praises the room but strongly criticizes check-in time in the same survey response” would not be the most direct first check for this failure. Whole-document sentiment can hide opposite opinions about different products, and literal polarity is unreliable for sarcasm.

C:

“The most powerful or least restricted model is always the safest choice regardless of task conditions” would not be the most direct first check for this failure. Whole-document sentiment can hide opposite opinions about different products, and literal polarity is unreliable for sarcasm.

D:

“Successful provisioning by itself proves all security, output, and operational constraints are satisfied” would not be the most direct first check for this failure. Whole-document sentiment can hide opposite opinions about different products, and literal polarity is unreliable for sarcasm.

Question 788

Situation: A hotel guest praises the room but strongly criticizes check-in time in the same survey response. Which result would demonstrate that the proposed remedy works end to end?

  1. A project dashboard remains green, although the affected request, output and permissions were not checked
  2. The operator reports success without reproducing the observed behavior: a hotel guest praises the room but strongly criticizes check-in time in the same survey response
  3. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages
  4. The team closes a change ticket without replaying a representative positive and negative test

Explanation

Correct Answer

C:

A hotel guest praises the room but strongly criticizes check-in time in the same survey response. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages. Evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions.

Incorrect Answers

A:

Because a hotel guest praises the room but strongly criticizes check-in time in the same survey response, “A project dashboard remains green, although the affected request, output and permissions were not checked” would not meet the requirement. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

B:

“The operator reports success without reproducing the observed behavior: a hotel guest praises the room but strongly criticizes check-in time in the same survey response” alone does not establish all the required behavior. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

D:

Because a hotel guest praises the room but strongly criticizes check-in time in the same survey response, “The team closes a change ticket without replaying a representative positive and negative test” alone does not establish all the required behavior. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

Question 789

Architecture case: A hotel guest praises the room but strongly criticizes check-in time in the same survey response. What evidence should the engineer correlate to verify the actual failure cause?

  1. Read only the broad service-health summary and skip actual workload evidence
  2. Restart every affected component before collecting request or tool traces
  3. Assume the general service is at fault without checking the affected operation in this case: a hotel guest praises the room but strongly criticizes check-in time in the same survey response
  4. Inspect aspect annotations, model confidence, sample errors, input language and disagreement among labelers

Explanation

Correct Answer

D:

A hotel guest praises the room but strongly criticizes check-in time in the same survey response. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages. Evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions. The observable acceptance condition here is: Inspect aspect annotations, model confidence, sample errors, input language and disagreement among labelers.

Incorrect Answers

A:

“Read only the broad service-health summary and skip actual workload evidence” would not meet the requirement. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

B:

“Restart every affected component before collecting request or tool traces” alone does not establish all the required behavior. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

C:

“Assume the general service is at fault without checking the affected operation in this case: a hotel guest praises the room but strongly criticizes check-in time in the same survey response” would not meet the requirement. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

Question 790

Production requirement: A hotel guest praises the room but strongly criticizes check-in time in the same survey response. Which TWO decisions would undermine the required system boundary? Select TWO answers.

  1. The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages
  2. Treat every positive adjective as overall customer satisfaction
  3. Evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions
  4. Escalate all neutral or mixed reviews as urgent complaints
  5. Inspect records that explain the affected process, then check the result described in this case: A hotel guest praises the room but strongly criticizes check-in time in the same survey response

Explanation

Correct Answer

B:

“Treat every positive adjective as overall customer satisfaction” would violate the requirement. The appropriate control instead is to evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions.

D:

“Escalate all neutral or mixed reviews as urgent complaints” would violate the requirement.

Incorrect Answers

A:

“The model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages” is the required outcome of a working control, not an action that weakens it.

C:

“Evaluate aspect-level sentiment using labeled examples and calibrate neutral and sarcasm cases before routing service actions” is the controlled implementation, not an unsafe workaround.

E:

“Inspect records that explain the affected process, then check the result described in this case: A hotel guest praises the room but strongly criticizes check-in time in the same survey response” adds a relevant safeguard or verification step rather than weakening the solution. Confirm that the model separates product, delivery and support aspects with calibrated labels on ambiguous and sarcastic validation messages.

Developing AI Apps and Agents on Azure (AI-103) Practice Tests