Custom Document Models Quiz

5 questions Pass: 70% +25 pts (50 with Pro)

Quiz covering Document Processing Fundamentals

What you'll be asked

  1. Which Azure AI Document Intelligence feature is best suited for extracting data from documents that have unique layouts or structures not covered by prebuilt models?
  2. When training a Custom Extraction Model, what is the minimum number of document samples typically recommended to achieve a reliable model?
  3. You are building a Custom Extraction Model and notice that the model consistently fails to extract a specific field that appears in different locations across your documents. What is the most effective way to improve performance?
  4. Which file format is NOT supported for training a Custom Document Model in Azure AI Document Intelligence?
  5. When building a Custom Neural Document Model, why might you choose to include 'negative' samples (documents that do not contain the fields you are trying to extract) in your training set?

Answers and explanations come after you submit.