Question 380
An operator can reproduce this error: An article portal shows generic word frequency rather than meaningful topics. Which targeted repair is justified?
- Apply entity extraction and review entity spans and categories in sample reports
- Identify sensitive spans and sanitize source and output under the retention policy
- Apply sentiment analysis with representative domain testing and interpret mixed cases carefully
- Use key phrase extraction and filter noisy or redundant phrases
Explanation
Correct Answer
D:The fault signal is that an article portal shows generic word frequency rather than meaningful topics; the targeted fix works because key phrase extraction returns salient terms or phrases; it differs from assigning one of several predefined category labels, restoring the ability to surface the main subjects of short articles without requiring a fixed taxonomy.
Incorrect Answers
A:If the failing function were to identify the people, organizations and places mentioned in incident narratives, this change could be relevant because named entity recognition identifies and classifies mentions in text; it does not by itself establish who authored the document or whether claims are true; the actual symptom is that an article portal shows generic word frequency rather than meaningful topics, which calls for a different repair.
B:If the failing function were to produce a useful incident synopsis without retaining unnecessary personal identifiers, this change could be relevant because a redaction stage can remove sensitive spans from the generated artifact; summarization alone does not guarantee privacy; the actual symptom is that an article portal shows generic word frequency rather than meaningful topics, which calls for a different repair.
C:If the failing function were to detect broadly positive, negative or neutral attitudes expressed in reviews, this change could be relevant because sentiment analysis estimates evaluative polarity in text; it should not be used as a precise diagnosis of the reviewer's emotions; the actual symptom is that an article portal shows generic word frequency rather than meaningful topics, which calls for a different repair.
Question 381
Stakeholders want a customer-facing AI service able to produce a useful incident synopsis without retaining unnecessary personal identifiers. Which choice implements that requirement?
- Detect sensitive personal information before reuse
- Generate a grounded document summary
- Redact private details from summarized text
- Summarize a conversation chronologically
Explanation
Correct Answer
C:For the stated requirement to produce a useful incident synopsis without retaining unnecessary personal identifiers, a redaction stage can remove sensitive spans from the generated artifact; summarization alone does not guarantee privacy; this is the mechanism that the system must implement.
Incorrect Answers
A:For a system whose objective is to find private identifiers in user-provided text so they can be minimized or masked, PII recognition identifies supported sensitive spans for redaction or controlled processing; keeping the raw text is a separate policy decision; this does not directly meet the required outcome to produce a useful incident synopsis without retaining unnecessary personal identifiers.
B:For a system whose objective is to condense a long policy document while preserving key exceptions and source support, grounded summarization uses supplied text as its evidence; a fluent summary must not introduce unsupported policy claims; this does not directly meet the required outcome to produce a useful incident synopsis without retaining unnecessary personal identifiers.
D:For a system whose objective is to create a handoff note that preserves commitments and unresolved action items from a long support call, conversation summaries must capture speakers, decisions and next steps rather than merely extract generic keywords; this does not directly meet the required outcome to produce a useful incident synopsis without retaining unnecessary personal identifiers.
Question 382
Engineers have implemented a change to classify messages against a defined taxonomy. What validation would expose a remaining defect?
- Measure entity-level precision and recall with annotated names and locations
- Review class-wise F1 and routing errors on a representative labeled set
- Measure valid-schema rate and field-level extraction accuracy on labeled requests
- Measure domain-category confusion on documents containing ambiguous terms
Explanation
Correct Answer
B:The accepted change is to classify messages against a defined taxonomy; this evaluation checks whether it can route written requests into one of several approved service categories, since a classifier assigns labels from an agreed taxonomy; named entity extraction alone does not decide message intent.
Incorrect Answers
A:This test concerns the ability to identify the people, organizations and places mentioned in incident narratives, for which named entity recognition identifies and classifies mentions in text; it does not by itself establish who authored the document or whether claims are true; it does not establish whether the change to classify messages against a defined taxonomy works for the current release.
C:This test concerns the ability to extract account ID, reason category, and decision text into fixed machine-consumable fields, for which structured output controls field names and types for downstream code; domain validation must still check allowed values; it does not establish whether the change to classify messages against a defined taxonomy works for the current release.
D:This test concerns the ability to extract a specialized compliance category when generic terminology is overloaded, for which task-specific definitions and examples help a model map ambiguous words to the intended domain meaning; it does not establish whether the change to classify messages against a defined taxonomy works for the current release.
Question 383
A release reviewer is assessing a feature intended to use domain examples to disambiguate terms. Which test is most meaningful?
- Review class-wise F1 and routing errors on a representative labeled set
- Measure domain-category confusion on documents containing ambiguous terms
- Measure valid-schema rate and field-level extraction accuracy on labeled requests
- Compare meaning preservation and style adherence using bilingual human review
Explanation
Correct Answer
B:The accepted change is to use domain examples to disambiguate terms; this evaluation checks whether it can extract a specialized compliance category when generic terminology is overloaded, since task-specific definitions and examples help a model map ambiguous words to the intended domain meaning.
Incorrect Answers
A:This test concerns the ability to route written requests into one of several approved service categories, for which a classifier assigns labels from an agreed taxonomy; named entity extraction alone does not decide message intent; it does not establish whether the change to use domain examples to disambiguate terms works for the current release.
C:This test concerns the ability to extract account ID, reason category, and decision text into fixed machine-consumable fields, for which structured output controls field names and types for downstream code; domain validation must still check allowed values; it does not establish whether the change to use domain examples to disambiguate terms works for the current release.
D:This test concerns the ability to translate a conversational response while preserving a specified tone and contextual meaning, for which a multilingual model can use conversation context and style constraints; factual meaning and fluent tone need separate verification; it does not establish whether the change to use domain examples to disambiguate terms works for the current release.
Question 384
The support ticket describes this behavior: A multilingual service assumes every incoming message is English and produces poor translations. Which engineering change should come first?
- Add an approved term mapping through supported translation or post-editing controls
- Detect probable language and route to a supported pipeline with uncertainty handling
- Call the supported text translation service for each target locale and evaluate results
- Apply PII detection and redact or tokenize the detected fields before logging
Explanation
Correct Answer
B:The fault signal is that a multilingual service assumes every incoming message is English and produces poor translations; the targeted fix works because language detection supplies a routing signal to a suitable translation or analysis path; short or mixed-language texts need caution, restoring the ability to handle incoming tickets whose language varies and is not declared by the sender.
Incorrect Answers
A:If the failing function were to preserve an approved glossary of medical or legal terms across translated responses, this change could be relevant because glossary or controlled terminology guidance reduces inconsistent domain-language choices; blind literal translation may be unsafe; the actual symptom is that a multilingual service assumes every incoming message is English and produces poor translations, which calls for a different repair.
C:If the failing function were to translate stable product instructions into multiple supported target languages, this change could be relevant because translator provides text translation with language configuration; domain terminology still requires quality review; the actual symptom is that a multilingual service assumes every incoming message is English and produces poor translations, which calls for a different repair.
D:If the failing function were to find private identifiers in user-provided text so they can be minimized or masked, this change could be relevant because PII recognition identifies supported sensitive spans for redaction or controlled processing; keeping the raw text is a separate policy decision; the actual symptom is that a multilingual service assumes every incoming message is English and produces poor translations, which calls for a different repair.
Question 385
The team plans to analyze sentiment in customer feedback. Which test provides the clearest evidence that this change works?
- Measure entity-level precision and recall with annotated names and locations
- Check sensitive-span recall and false-positive removal on realistic tickets
- Measure agreement with human sentiment labels, including mixed-polarity reviews
- Compare extracted topics with editor-labeled key concepts across articles
- Evaluate language identification accuracy on short and code-switched messages
Explanation
Correct Answer
C:The accepted change is to analyze sentiment in customer feedback; this evaluation checks whether it can detect broadly positive, negative or neutral attitudes expressed in reviews, since sentiment analysis estimates evaluative polarity in text; it should not be used as a precise diagnosis of the reviewer's emotions.
Incorrect Answers
A:This test concerns the ability to identify the people, organizations and places mentioned in incident narratives, for which named entity recognition identifies and classifies mentions in text; it does not by itself establish who authored the document or whether claims are true; it does not establish whether the change to analyze sentiment in customer feedback works for the current release.
B:This test concerns the ability to find private identifiers in user-provided text so they can be minimized or masked, for which PII recognition identifies supported sensitive spans for redaction or controlled processing; keeping the raw text is a separate policy decision; it does not establish whether the change to analyze sentiment in customer feedback works for the current release.
D:This test concerns the ability to surface the main subjects of short articles without requiring a fixed taxonomy, for which key phrase extraction returns salient terms or phrases; it differs from assigning one of several predefined category labels; it does not establish whether the change to analyze sentiment in customer feedback works for the current release.
E:This test concerns the ability to handle incoming tickets whose language varies and is not declared by the sender, for which language detection supplies a routing signal to a suitable translation or analysis path; short or mixed-language texts need caution; it does not establish whether the change to analyze sentiment in customer feedback works for the current release.
Question 386
Production telemetry reveals this failure: A chatbot logs inappropriate user content with no review or filtering signal. Which change most directly corrects it?
- Use a source-constrained summary prompt and validate coverage of required clauses
- Extract opinion targets and their associated sentiment rather than only overall polarity
- Apply supported text content-safety checks and enforce category-specific handling
- Apply structured conversation summarization with action-item and attribution requirements
Explanation
Correct Answer
C:The fault signal is that a chatbot logs inappropriate user content with no review or filtering signal; the targeted fix works because text safety classification identifies supported content categories; moderation decisions depend on configured policy and context, restoring the ability to review user messages for policy-relevant hate, violence or self-harm content before generation.
Incorrect Answers
A:If the failing function were to condense a long policy document while preserving key exceptions and source support, this change could be relevant because grounded summarization uses supplied text as its evidence; a fluent summary must not introduce unsupported policy claims; the actual symptom is that a chatbot logs inappropriate user content with no review or filtering signal, which calls for a different repair.
B:If the failing function were to identify that a reviewer likes delivery speed but dislikes package quality in one comment, this change could be relevant because opinion mining links sentiment to specific targets or aspects instead of reporting only a document-level score; the actual symptom is that a chatbot logs inappropriate user content with no review or filtering signal, which calls for a different repair.
D:If the failing function were to create a handoff note that preserves commitments and unresolved action items from a long support call, this change could be relevant because conversation summaries must capture speakers, decisions and next steps rather than merely extract generic keywords; the actual symptom is that a chatbot logs inappropriate user content with no review or filtering signal, which calls for a different repair.
Question 387
The acceptance criteria for a customer-facing AI service require it to translate a conversational response while preserving a specified tone and contextual meaning. Which design should be approved?
- Use Azure Translator for text localization
- Apply terminology constraints to regulated translations
- Redact private details from summarized text
- Use LLM translation with contextual instructions
Explanation
Correct Answer
D:For the stated requirement to translate a conversational response while preserving a specified tone and contextual meaning, a multilingual model can use conversation context and style constraints; factual meaning and fluent tone need separate verification; this is the mechanism that the system must implement.
Incorrect Answers
A:For a system whose objective is to translate stable product instructions into multiple supported target languages, translator provides text translation with language configuration; domain terminology still requires quality review; this does not directly meet the required outcome to translate a conversational response while preserving a specified tone and contextual meaning.
B:For a system whose objective is to preserve an approved glossary of medical or legal terms across translated responses, glossary or controlled terminology guidance reduces inconsistent domain-language choices; blind literal translation may be unsafe; this does not directly meet the required outcome to translate a conversational response while preserving a specified tone and contextual meaning.
C:For a system whose objective is to produce a useful incident synopsis without retaining unnecessary personal identifiers, a redaction stage can remove sensitive spans from the generated artifact; summarization alone does not guarantee privacy; this does not directly meet the required outcome to translate a conversational response while preserving a specified tone and contextual meaning.
Question 388
A design review proposes to apply terminology constraints to regulated translations. Which measurable check would establish success on real inputs?
- Use human adequacy and terminology checks for representative translated instructions
- Evaluate language identification accuracy on short and code-switched messages
- Measure terminology consistency and reviewer corrections across languages
- Review class-wise F1 and routing errors on a representative labeled set
Explanation
Correct Answer
C:The accepted change is to apply terminology constraints to regulated translations; this evaluation checks whether it can preserve an approved glossary of medical or legal terms across translated responses, since glossary or controlled terminology guidance reduces inconsistent domain-language choices; blind literal translation may be unsafe.
Incorrect Answers
A:This test concerns the ability to translate stable product instructions into multiple supported target languages, for which translator provides text translation with language configuration; domain terminology still requires quality review; it does not establish whether the change to apply terminology constraints to regulated translations works for the current release.
B:This test concerns the ability to handle incoming tickets whose language varies and is not declared by the sender, for which language detection supplies a routing signal to a suitable translation or analysis path; short or mixed-language texts need caution; it does not establish whether the change to apply terminology constraints to regulated translations works for the current release.
D:This test concerns the ability to route written requests into one of several approved service categories, for which a classifier assigns labels from an agreed taxonomy; named entity extraction alone does not decide message intent; it does not establish whether the change to apply terminology constraints to regulated translations works for the current release.
Question 389
An Azure AI delivery has 2 independently testable requirements: (1) surface the main subjects of short articles without requiring a fixed taxonomy; and (2) identify that a reviewer likes delivery speed but dislikes package quality in one comment. Select the most directly specialized implementation for each requirement. Choose TWO.
- Extract key phrases for topic browsing
- Use opinion mining for aspect-level sentiment
- Extract named entities from reports
- Use Azure Translator for text localization
- Analyze sentiment in customer feedback
Explanation
Correct Answers
A:For the stated requirement to surface the main subjects of short articles without requiring a fixed taxonomy, key phrase extraction returns salient terms or phrases; it differs from assigning one of several predefined category labels; this is the mechanism that the system must implement. This choice satisfies the specific deliverable to surface the main subjects of short articles without requiring a fixed taxonomy, while the other deliverables require separate capabilities.
B:For the stated requirement to identify that a reviewer likes delivery speed but dislikes package quality in one comment, opinion mining links sentiment to specific targets or aspects instead of reporting only a document-level score; this is the mechanism that the system must implement. This choice satisfies the specific deliverable to identify that a reviewer likes delivery speed but dislikes package quality in one comment, while the other deliverables require separate capabilities.
Incorrect Answers
C:For a system whose objective is to identify the people, organizations and places mentioned in incident narratives, named entity recognition identifies and classifies mentions in text; it does not by itself establish who authored the document or whether claims are true; this does not directly meet the required outcome to identify that a reviewer likes delivery speed but dislikes package quality in one comment.
D:For a system whose objective is to translate stable product instructions into multiple supported target languages, translator provides text translation with language configuration; domain terminology still requires quality review; this does not directly meet the required outcome to identify that a reviewer likes delivery speed but dislikes package quality in one comment.
E:For a system whose objective is to detect broadly positive, negative or neutral attitudes expressed in reviews, sentiment analysis estimates evaluative polarity in text; it should not be used as a precise diagnosis of the reviewer's emotions; this does not directly meet the required outcome to identify that a reviewer likes delivery speed but dislikes package quality in one comment.
Question 390
A user-acceptance test fails with the following observation: A literal translator preserves words but loses a required conversational form of address in a nuanced reply. What is the best correction?
- Translate with context and explicit tone requirements, then assess semantic fidelity
- Detect probable language and route to a supported pipeline with uncertainty handling
- Use a source-constrained summary prompt and validate coverage of required clauses
- Call the supported text translation service for each target locale and evaluate results
Explanation
Correct Answer
A:The fault signal is that a literal translator preserves words but loses a required conversational form of address in a nuanced reply; the targeted fix works because a multilingual model can use conversation context and style constraints; factual meaning and fluent tone need separate verification, restoring the ability to translate a conversational response while preserving a specified tone and contextual meaning.
Incorrect Answers
B:If the failing function were to handle incoming tickets whose language varies and is not declared by the sender, this change could be relevant because language detection supplies a routing signal to a suitable translation or analysis path; short or mixed-language texts need caution; the actual symptom is that a literal translator preserves words but loses a required conversational form of address in a nuanced reply, which calls for a different repair.
C:If the failing function were to condense a long policy document while preserving key exceptions and source support, this change could be relevant because grounded summarization uses supplied text as its evidence; a fluent summary must not introduce unsupported policy claims; the actual symptom is that a literal translator preserves words but loses a required conversational form of address in a nuanced reply, which calls for a different repair.
D:If the failing function were to translate stable product instructions into multiple supported target languages, this change could be relevant because translator provides text translation with language configuration; domain terminology still requires quality review; the actual symptom is that a literal translator preserves words but loses a required conversational form of address in a nuanced reply, which calls for a different repair.