Extract Information with Azure Content Understanding in Foundry Tools AI-102 Practice Questions (Page 9)

Questions 41–45

1. 41expert · hard

   A retail company receives supplier invoices that contain a table of line items, a subtotal, taxes, and a total. They need to extract the line items, the invoice number, and the total amount. The invoices are in various formats, some are scanned, and some are digital PDFs. They need high accuracy and want to handle variations in table layout. What should they do in Azure Content Understanding?

   • AUse an entity extraction pipeline to extract all numbers and then manually map them to the invoice number and total.
   • BCreate a custom table extraction pipeline with OCR enabled, and add custom fields for invoice number and total amount, then test with a diverse set of invoices.
   • CUse a pre-built table extraction model and assume it will handle all invoice formats without customization.
   • DUse a document classification pipeline to categorize the invoice and then use the category to infer the line items and total.
   Reveal Solution Discussion Select an answer first
2. 42foundation · easy

   What is the main purpose of image extraction in Azure Content Understanding?

   • ATo extract text from images
   • BTo summarize documents
   • CTo classify documents
   • DTo extract images from documents for further analysis or processing
   Reveal Solution Discussion Select an answer first
4. 43application · medium

   A news agency receives press releases in various formats. They need to automatically classify each press release as 'urgent', 'standard', or 'low priority', and also generate a brief summary for their editorial team. They also want to extract the date of the event mentioned. What should they implement in Azure Content Understanding?

   • AA pipeline that only extracts entities and then uses the entity types to determine the priority.
   • BA pipeline that only classifies the press release and uses the classification label as the summary.
   • CA pipeline that classifies the press release, generates a summary, and extracts the event date as a custom field.
   • DA pipeline that only summarizes the press release and then manually assigns the priority based on the summary.
   Reveal Solution Discussion Select an answer first
5. 44foundation · easy

   Which Azure Content Understanding feature would you use to convert a table in a scanned report into a structured format like CSV?

   • ATable extraction
   • BOCR pipeline
   • CEntity extraction
   • DDocument classification
   Reveal Solution Discussion Select an answer first
6. 45application · medium

   A logistics company needs to digitize incoming bills of lading. Each document contains a mix of printed text, a small table of item quantities, and a barcode. The team wants to extract the barcode value, the table rows, and the overall text using Azure Content Understanding in Foundry Tools. They plan to create a single pipeline that handles all these elements. What should they configure?

   • ACreate a table extraction pipeline and rely on the table output to include the barcode as a cell.
   • BCreate a custom extraction pipeline with OCR enabled, then add a table extraction field and a barcode attribute field.
   • CCreate a summarization pipeline that generates a summary containing the barcode and table data.
   • DCreate a standard OCR pipeline and then use a separate document classification model to identify the barcode.
   Reveal Solution Discussion Select an answer first