Question 39
Domain 4: Implement Natural Language Processing SolutionsA travel booking bot uses CLU and correctly identifies the `BookFlight` intent but fails to extract the departure city when users say "I need to fly out of JFK." The entity type for airports is defined as an ML entity trained on common formats. What should you do?
Correct answer: A
Explanation
A List entity is used for exact, known values, so adding "JFK" and similar airport codes gives the bot a reliable match for those variations. An ML entity trained on common formats may miss shorthand or uncommon airport references, so supplementing it with a list improves extraction for phrases like "fly out of JFK."
Why each option is right or wrong
A. Add `JFK` and similar variations to a List entity for airports
Under Azure AI Language Conversational Language Understanding, an ML entity is learned from examples and generalizes patterns, but it is not guaranteed to catch short airport codes like “JFK” unless those variants are represented in training data. A List entity is the correct fit for fixed, exact-value terms such as airport codes; adding “JFK” and other known codes gives deterministic matching for those 3-letter values, which is why the departure city can then be extracted from phrases like “fly out of JFK.”
B. Change the ML entity to a Prebuilt entity
C. Increase the training utterance count to 100
D. Enable entity linking in Azure AI Language