AI-102 Question 389
Single answerYour organization handles thousands of invoices monthly and wants to automate the extraction of key information such as invoice numbers, dates, and total amounts. You decide to use Azure AI Document Intelligence (formerly Form Recognizer). After setting up the service, what is the next step to customize the solution to extract this specific information?
- A
Use the pre-built invoice model for invoice data extraction.
- B
Create and train a custom model using labeled data specific to your invoices.
- C
Use the OCR API to extract text and manually parse the required information.
- D
Deploy an Azure Cognitive Search index to process and extract invoice data.
Show answer and explanation
Correct answer: B
Explanation
To customize the solution for extracting specific information from invoices, a custom model should be created and trained with labeled examples of your invoice formats. This ensures accurate extraction of fields unique to your organization's requirements. While the pre-built invoice model is useful for general cases, it may not handle unique customizations effectively.
- A. Incorrect.
The pre-built invoice model can be used for extracting common fields from invoices, but it may not meet the customization needs for specific invoice formats unique to your business.
- B. Correct.
Creating and training a custom model using labeled data specific to your invoices allows you to tailor the extraction process to handle unique formats and required fields.
- C. Incorrect.
Using the OCR API can extract raw text, but it lacks the intelligence to identify and structure specific fields such as invoice numbers or dates effectively.
- D. Incorrect.
Azure Cognitive Search is used for indexing and searching data, not for extracting structured information from documents.