Analyze and Translate Text AI-102 Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A multinational HR department collects employee feedback in multiple languages. They need to extract main topics from each comment, detect the language for reporting, and redact any personal data before storing. The solution must minimize API calls and ensure that no PII is stored. The team is considering two pipelines: (A) redact PII first, then detect language and extract key phrases; (B) detect language and extract key phrases first, then redact PII from the results. Which pipeline should you choose and why?

   • APipeline A, but only if the text is in English, because redaction may remove characters needed for language detection.
   • BPipeline A, because redacting PII first ensures that no PII is stored and subsequent analyses operate on safe text.
   • CPipeline B, because key phrase extraction requires the original text to identify topics, and redaction would remove key phrases.
   • DPipeline B, because language detection and key phrase extraction are more accurate on the original text, and redacting the results is sufficient.
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A multilingual support chatbot logs user messages. The team wants to detect the language of each message, redact any personal data before logging, and determine the sentiment to escalate negative interactions. The solution must use Azure AI services and ensure that no PII is stored in logs. What should you do?

   • ACall Sentiment Analysis only, because it returns both language and PII entities.
   • BUse the Translator API to translate messages to English, then run PII Detection and Sentiment Analysis on the translated text.
   • CCall Language Detection and Sentiment Analysis on the original text, then redact PII from the results before logging.
   • DCall PII Detection to redact PII, then call Language Detection and Sentiment Analysis on the redacted text.
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A legal firm processes contracts in multiple languages. They need to extract the main topics from each contract and also identify the parties involved (people, organizations). The solution must minimize the number of API calls and work on the original text without translation. What should you do?

   • ATranslate the contract to English first, then call Key Phrase Extraction and Entity Recognition on the translated text.
   • BUse the Translator API with a custom dictionary to extract key phrases and entities.
   • CCall Entity Recognition only, because it returns both key phrases and entities in a single response.
   • DCall Key Phrase Extraction to get main topics, then call Entity Recognition to identify people and organizations from the same text.
   Reveal Solution Discussion Select an answer first
5. 4expert · hard

   A social media analytics company processes posts that may contain personal data. The company must determine sentiment of each post, detect the language for demographic analysis, and ensure that any email addresses are masked before storing. The solution must minimize the risk of exposing PII during processing. You need to design the pipeline. What should you do?

   • ARun PII Detection to mask email addresses first, then run Sentiment Analysis and Language Detection on the masked text.
   • BRun Language Detection first, then Sentiment Analysis on the original text, and then PII Detection to mask email addresses.
   • CRun Sentiment Analysis on the original text, then PII Detection to mask email addresses, and then Language Detection on the masked text.
   • DRun Sentiment Analysis and Language Detection on the original text, then run PII Detection to mask email addresses.
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   A multilingual news aggregator receives articles in various languages. The team wants to detect the language of each article, translate non-English articles to English, and extract key phrases from the translated article for tagging. The solution must minimize API calls and ensure that key phrases are relevant to the English version. The team is considering two approaches: (A) call Language Detection, then the Translator API, then Key Phrase Extraction on the translated text; (B) call the Translator API, which returns the detected source language, then Key Phrase Extraction on the translated text. Which approach should you choose and why?

   • AApproach A, because Language Detection is more accurate than the Translator API's language detection.
   • BApproach A, because Key Phrase Extraction should be run on the original text to preserve the original topics.
   • CApproach B, but only if the article is in a language supported by the Translator API; otherwise, use Language Detection.
   • DApproach B, because the Translator API returns the detected source language, so a separate Language Detection call is unnecessary, and key phrases are extracted from the translated text.
   Reveal Solution Discussion Select an answer first