AI-102 Implement knowledge mining and information extraction solutions Practice Test 8 — 15 Questions

Question 1

You are designing a knowledge mining solution for a legal firm. The solution must index contracts and support queries like 'find all contracts where the governing law is New York'. The contracts are stored as PDFs in Azure Blob Storage. Which approach should you use?

Accepted Answer

Use Azure AI Search with a custom skill to extract the governing law field. Option C is correct because Azure AI Search with a custom skill to extract the governing law field allows indexing that specific metadata. Option A is wrong because full-text search on raw text would not distinguish the governing law field. Option B is wrong because skillsets can extract fields, not just enrich. Option D is wrong because vector search is for semantic similarity, not exact field extraction.

Answer

Use Azure AI Search with vector search

Answer

Use Azure AI Search with a built-in skillset for OCR

Answer

Use Azure AI Search with full-text search on the extracted text

Question 2

You are using Azure AI Document Intelligence to extract data from purchase orders. The purchase orders have a table of line items. Which prebuilt model should you use?

Accepted Answer

Prebuilt invoice model. Option B is correct because the prebuilt invoice model extracts tables and line items. Option A is wrong because the general document model extracts text and layout but not invoice-specific fields. Option C is wrong because the custom model requires training data. Option D is wrong because the layout model only extracts text and tables without field mapping.

Answer

Prebuilt document model

Answer

Layout model

Answer

Custom model

Question 3

Your knowledge mining solution uses Azure AI Search with cognitive skills. During testing, you notice that some documents are not being enriched because the skillset execution fails. Which diagnostic step should you take first?

Accepted Answer

Enable debug mode on the skillset. Option B is correct because enabling debug mode creates a debug session that captures detailed execution logs for each skill. Option A is wrong because checking the indexer status shows success/failure but not detailed skill errors. Option C is wrong because resubmitting might repeat the same error without insight. Option D is wrong because the portal shows indexer execution history but not per-document skill details.

Answer

Review the indexer execution history in the portal

Answer

Re-run the indexer with a fresh document

Answer

Check the indexer logs in Azure Monitor

Question 4

You need to build a chatbot that answers questions based on your company's internal knowledge base. The knowledge base consists of Word documents and PDFs. Which service should you use to create a conversational interface that retrieves answers from these documents?

Accepted Answer

Azure AI Search with Azure AI Bot Service. Option D is correct because Azure AI Search can index documents and be used as a data source for a question-answering system, and Azure AI Bot Service can host the chatbot. Option A is wrong because Language Service's QnA Maker (now Custom Question Answering) can use a knowledge base but not directly index documents; it requires a search index for large datasets. Option B is wrong because Document Intelligence extracts data but does not answer questions. Option C is wrong because Computer Vision does not handle text queries.

Answer

Azure AI Language Service - Custom Question Answering

Answer

Azure AI Computer Vision

Answer

Azure AI Document Intelligence

Question 5

You are designing a knowledge mining solution that must handle sensitive customer data. The solution must ensure that personally identifiable information (PII) is not returned in search results. What should you do?

Accepted Answer

Use a custom skill in the skillset to detect and redact PII before indexing. Option B is correct because enabling PII detection in the enrichment pipeline and using a custom skill to remove or mask PII fields before indexing prevents PII from being stored in the index. Option A is wrong because encryption does not prevent PII from being returned in results. Option C is wrong because field mappings control how fields are imported, not content removal. Option D is wrong because access control restricts who can search but does not remove PII from results.

Answer

Use Azure AI Search with encryption at rest

Answer

Implement role-based access control on the search index

Answer

Configure field mappings to exclude PII fields

Question 6

You are extracting text from scanned documents that are in French. Which capability of Azure AI Document Intelligence should you use?

Accepted Answer

Read API. Option C is correct because the Read API supports multiple languages including French for OCR. Option A is wrong because the Layout model extracts tables but not necessarily text in specific languages. Option B is wrong because the custom model requires training. Option D is wrong because the prebuilt invoice model is for invoices and may not support all languages or general text extraction.

Answer

Custom model

Answer

Layout model

Answer

Prebuilt invoice model

Question 7

Which TWO services can be used to enrich an Azure AI Search index with knowledge mining skills? (Choose two.)

Accepted Answer

Azure AI Computer Vision. Options B and D are correct. Azure AI Language Service provides entity recognition, key phrase extraction, etc. Azure AI Computer Vision provides image analysis, OCR, etc. Option A is wrong because Translator is for translation, not enrichment. Option C is wrong because Speech Service is for audio. Option E is wrong because Video Indexer is for video analysis.

Answer

Azure AI Video Indexer

Answer

Azure AI Speech Service

Answer

Azure AI Translator

Question 8

Which THREE actions should you take when designing a custom skill for an Azure AI Search enrichment pipeline? (Choose three.)

Accepted Answer

Define input and output parameters in the skillset definition. Options A, C, and D are correct. The skill must be deployed as a web API that accepts JSON input and returns JSON output. It should handle errors gracefully. It must be registered in the skillset. Option B is wrong because the skill should not be long-running; indexer timeouts apply. Option E is wrong because the skill can be in any language as long as it exposes a REST API.

Answer

Ensure the skill executes within 2 minutes

Answer

Write the skill in a language supported by Azure AI Search

Question 9

Which TWO capabilities are available in Azure AI Search to improve search relevance? (Choose two.)

Accepted Answer

Scoring profiles. Options A and D are correct. Scoring profiles allow boosting by field values or freshness. Semantic ranking re-ranks results to improve relevance. Option B is wrong because synonym maps improve recall, not relevance. Option C is wrong because filters restrict results but do not improve relevance ranking. Option E is wrong because indexers are for data ingestion, not relevance.

Answer

Filters

Answer

Indexers

Answer

Synonym maps

Question 10

Refer to the exhibit. You have this skillset definition for an Azure AI Search enrichment pipeline. You notice that the entity recognition skill is not executing on any document. What is the most likely cause?

Accepted Answer

The input source path in the entity recognition skill should be relative to the context, not absolute. Option C is correct. The context of the EntityRecognitionSkill is '/document/pages/*', but the SplitSkill outputs 'pages' at '/document/pages'. However, the SplitSkill's output is named 'pages' but the context for the split skill is '/document', so the output path is '/document/pages'. The entity recognition skill context '/document/pages/*' would iterate over each element in '/document/pages', but if the split skill's output is not an array, the iteration fails. Actually, the split skill outputs 'textItems' as an array, but the target name is 'pages', so the output is at '/document/pages', which is an array. The context '/document/pages/*' should work. However, the issue is that the split skill's output is 'textItems' but the target name is 'pages', so the actual output node is '/document/pages'. The entity recognition skill inputs source '/document/pages/*' expects each page's content, but the input field name is 'text' and source is correct. Another potential issue is that the split skill's output is not being passed correctly because the entity recognition skill is not referencing the correct output from the split skill. Actually, the correct output should be '/document/pages' which is an array of strings. The entity recognition skill context '/document/pages/*' should iterate over each page. But the entity recognition skill's input source is '/document/pages/*', which is incorrect because that would be the element itself, not the text content. The input source should be '/document/pages/*' to get the text of each page. However, the split skill outputs the text items as strings, so '/document/pages/*' would be the string content. That should work. Wait, the exhibit shows the split skill output target name 'pages', so the output node is '/document/pages' (array). The entity recognition skill input source is '/document/pages/*', which is the individual page string. That seems correct. However, the entity recognition skill expects a 'text' input, and the source is '/document/pages/*' which is the page text. So why would it not execute? Possibly because the language code source '/document/language' is not present in the document. But that would cause an error for the split skill too. Another reason could be that the entity recognition skill requires the language code to be provided, and if it's missing, the skill fails. But the question says 'not executing on any document', implying it never runs. The most likely cause is that the split skill is not producing pages because the maximumPageLength might be too large and the content is short, but that would still produce one page. Actually, the split skill will always produce at least one page. The exhibit shows the skillset, but the entity recognition skill context is '/document/pages/*' which is correct. However, the entity recognition skill might fail if the language code is invalid. But the most common mistake is that the entity recognition skill's context is set to '/document/pages/*', but the input source is '/document/pages/*', which is the same as context, leading to no iteration. Actually, the context defines the iteration, and the input source should be the property of that context, not the context itself. For example, if context is '/document/pages/*', then input source should be relative to that context, like 'text' if the page object had a 'text' property. But here, the input source is '/document/pages/*' which is an absolute path that points to the same node as the context, so it might cause a conflict. In Azure AI Search, the input source should be a path relative to the context or absolute. If the source is absolute and points to the same node as the context, it might not work as expected because the skill expects the input to be a scalar value, but the context is an array element. This is a known issue. Option C states that the input source path should be relative to the context, not absolute. That is the correct answer.

Answer

The entity recognition skill requires a language code that is missing

Answer

The split skill is not producing pages because the content is too short

Answer

The entity recognition skill is not registered in the skillset

Question 11

Refer to the exhibit. You have this Azure AI Search indexer configuration. The indexer is failing after processing 6 documents that contain errors. What should you do to ensure the indexer continues processing even if some documents fail?

Accepted Answer

Increase maxFailedItems to a higher value, such as 100. Option B is correct. The current configuration has maxFailedItems=5, meaning the indexer stops after 5 failures. Increasing maxFailedItems to a higher value (e.g., 100) allows the indexer to continue. Option A is wrong because decreasing batch size may reduce failures but does not increase the failure tolerance. Option C is wrong because removing the schedule does not affect failure handling. Option D is wrong because increasing batch size may increase failures.

Answer

Decrease the batch size to 5

Answer

Increase batch size to 20

Answer

Remove the schedule to run the indexer on demand

Question 12

Refer to the exhibit. You execute a search query on an Azure AI Search index and get these results. The query was 'brown fox'. Why is the first result scored higher than the second?

Accepted Answer

The first document has a higher term frequency and better term proximity for the query terms. Option A is correct. The default scoring algorithm uses TF-IDF, which gives higher scores to documents where the query terms appear more frequently. The first document contains 'brown' and 'fox' exactly, while the second uses 'brown' and 'fox' but also has 'fast' and 'leaps', but the exact phrase match is stronger. Actually, the first document has the exact phrase 'brown fox' while the second has 'brown fox' separated. The first document likely has a higher term frequency or better proximity. Option B is wrong because there is no semantic ranking configured. Option C is wrong because there is no scoring profile defined. Option D is wrong because the index is not using vector search.

Answer

The first document has a higher value in a scoring profile field

Answer

The first document is more similar to the query in vector space

Answer

The first document was boosted by a semantic ranking function

Question 13

You are building a solution to extract key information from invoices using Azure AI Document Intelligence. The invoices contain fields such as invoice number, date, total amount, and line items. However, the model is not correctly extracting the line items. Which prebuilt model should you use?

Accepted Answer

Prebuilt-invoice model. Option B is correct because the prebuilt-invoice model is specifically designed to extract fields from invoices, including line items. Option A is wrong because the prebuilt-layout model extracts text and structure but not specific key-value pairs like line items. Option C is wrong because the prebuilt-receipt model is for receipts, not invoices. Option D is wrong because the prebuilt-idDocument model is for identity documents.

Answer

Prebuilt-receipt model

Answer

Prebuilt-idDocument model

Answer

Prebuilt-layout model

Question 14

Your company uses Azure Cognitive Search to index millions of documents. Users report that search results include irrelevant documents. You need to improve search relevance by boosting documents that contain the search term in the title field. Which scoring profile configuration should you use?

Accepted Answer

Create a tagging scoring profile that boosts by the title field with a weight of 10.. Option A is correct because a tagging profile with a weight boosts documents based on matching tags from a specific field. Option B is wrong because a distance profile is for geospatial scoring. Option C is wrong because a magnitude profile is for boosting based on numeric values. Option D is wrong because a freshness profile boosts based on date/time.

Answer

Create a freshness scoring profile with a boosting duration of 30 days.

Answer

Create a distance scoring profile with a reference point parameter.

Answer

Create a magnitude scoring profile with a boosting function of 'linear'.

Question 15

You are using Azure AI Language to perform entity recognition on customer feedback. You need to identify the sentiment expressed towards specific entities. Which feature should you use?

Accepted Answer

Sentiment analysis with opinion mining. Option C is correct because sentiment analysis with opinion mining provides sentiment at the aspect level, including entities. Option A is wrong because Named Entity Recognition (NER) only identifies entities, not sentiment. Option B is wrong because key phrase extraction only extracts key phrases. Option D is wrong because entity linking connects entities to knowledge bases, not sentiment analysis.

Answer

Named Entity Recognition (NER)

Answer

Entity linking

Answer

Key phrase extraction