Implement Custom Language Models AI-102 Practice Questions (Page 14)

Questions 66–70

1. 66foundation · easy

   In the context of building a custom language model for a conversational application, what is the purpose of defining an intent?

   • ATo capture specific data points such as dates, numbers, or names from user input
   • BTo represent the action or goal that the user wants to accomplish
   • CTo provide sample phrases that help train the model to recognize variations of user input
   • DTo define the response that the application should give after recognizing the user's input
   Reveal Solution Discussion Select an answer first
2. 67foundation · easy

   What is the role of chit-chat in a question answering knowledge base?

   • ATo create multi-turn conversation flows
   • BTo improve the accuracy of entity extraction
   • CTo handle casual or small-talk queries that are not related to the main domain
   • DTo translate the knowledge base into multiple languages
   Reveal Solution Discussion Select an answer first
4. 68application · medium

   A company needs to translate internal documents from English to Japanese. The standard Azure Translator service does not perform well on the company's technical terminology. They have a parallel corpus of English-Japanese documents. What should they do to improve translation quality?

   • AUse Azure AI Language's custom question answering to translate the documents.
   • BUse the standard Translator service with a custom glossary file.
   • CTrain a custom translator model using the parallel corpus and publish it for use.
   • DManually translate all documents using a professional translation service.
   Reveal Solution Discussion Select an answer first
5. 69foundation · easy

   What is the purpose of training a knowledge base in Azure AI Language question answering?

   • ATo export the knowledge base to a file
   • BTo add more external sources
   • CTo deploy the knowledge base to a production endpoint
   • DTo improve the accuracy of the answers returned for user queries
   Reveal Solution Discussion Select an answer first
6. 70foundation · easy

   Which of the following is a common technique to optimize a custom language model's accuracy?

   • AChanging the model's name
   • BAdding more diverse sample utterances for each intent
   • CIncreasing the number of deployments
   • DDeleting the model and starting over
   Reveal Solution Discussion Select an answer first