AWS Application Services

Question 1

Using pre-defined logic and rules to make product recommendations to online shoppers is an example of machine learning.

Answer 1

False

Question 2

Which level of the ML stack helps you build custom ML models without managing infrastructure?

Answer 2

Middle level (Amazon SageMaker)

Question 3

Which of the following can you do with Amazon Textract?
a. Detect key-value pairs in documents

b. Build a custom ML model for text extraction
c. Send text extraction with low confidence scores for human review
d. Translate the detected text into English

Answer 3

a. Detect key-value pairs in documents

c. Send text extraction with low confidence scores for human review

Question 4

With Amazon Comprehend, a new model can be trained to help you extract custom entities from text.

Answer 4

True

Question 5

Which of the following is an example of the type of data Amazon Comprehend is designed to analyze?
a. Social media posts

b. Data in a table
c. Log files
d. GPS data

Answer 5

a. Social media posts

Question 6

When calling the DetectKeyPhrases API, which of the following is not returned by Amazon Comprehend?
a. The key phrases

b. The count of each key phrase
c. The confidence level for each key phrase
d. The sentiment of each key phrase

Answer 6

d. The sentiment of each key phrase

Question 7

You want to create a Lex bot that can help you order pizza. Why is it important to add slots as part of intent configuration?
a. So you can customize your orders with different pizza sizes and toppings.

b. So you can account for different ways you might convey your intent to order pizza.
c. So that a lambda function can be automatically set up for you to fulfill the intent.

Answer 7

a. So you can customize your orders with different pizza sizes and toppings.

Question 8

Let’s say you’re responsible for building a system that analyzes the sentiment of a customer chat. Which service should you integrate with Amazon Lex to do this?

Answer 8

Amazon Comprehend

Question 9

In which situation would an Amazon Lex fallback intent help?
a. When a user orders pizza but, due to background noise, the bot needs the user to repeat what they said.

b. When a bot has to use a previous exchange with a user to pretend to understand an unclear message from that user.
c. When a bot is asked a question that is not programmed to answer.

Answer 9

When a bot is asked a question that is not programmed to answer.

Question 10

Which three of the following options can Amazon Textract handle that traditional OCR methods are not able to?
a. Extracting words and lines from documents

b. Extracting forms (key/values) from documents without using any templates
c. Handling non-textual content such as radio buttons and checkboxes
d. Preserving the composition of data stored in tables

Answer 10

b. Extracting forms (key/values) from documents without using any templates
c. Handling non-textual content such as radio buttons and checkboxes
d. Preserving the composition of data stored in tables

Question 11

Which of the following is a common use case for integrating Amazon Textract with Amazon A2I (human review)?
a. You want to identify form labels and values from an image or PDF document.

b. You are extracting data from a document that requires review due to regulatory requirements or sensitive business decisions.
c. You have tables in your document, and you need to preserve the composition of the data stored in those tables.

Answer 11

b. You are extracting data from a document that requires review due to regulatory requirements or sensitive business decisions.

Question 12

You are looking to extract form or table data in a document. You need to do this synchronously because your use is latency-sensitive, such as mobile capture. What API should you use?
a. AnalyzeDocument

b. DetectDocumentText
c. StartDocumentAnalysis
d. GetDocumentTextDetection

Answer 12

a. AnalyzeDocument