AI-102 Question 291
Select 3You are tasked with implementing a custom question answering solution for an internal knowledge base using Azure AI Language. The knowledge base consists of product documentation and frequently asked questions (FAQs). You want to ensure that users can query the knowledge base in natural language and receive accurate responses. Which steps should you perform to create and deploy the solution?
- A
Use the Azure AI Language Studio to create a custom question answering project and upload your knowledge base documents.
- B
Train a custom language model using Azure Machine Learning and integrate it with the question answering project.
- C
Publish the custom question answering project to an endpoint for integration with client applications.
- D
Test the deployed endpoint by querying it with sample questions to validate the accuracy of the responses.
- E
Enable the Translation feature in Azure AI Language to support multilingual queries automatically.
Show answer and explanation
Correct answers: A, C, D
Explanation
To create a custom question answering solution using Azure AI Language, you must first create a project in Azure AI Language Studio and upload your knowledge base documents. After configuring and training the model, you need to publish the project to an endpoint for client integration. Finally, testing the endpoint with sample queries ensures the solution works as intended. Training a custom language model or enabling unrelated features like Translation is unnecessary for this specific task.
- A. Correct.
Correct: Azure AI Language Studio provides tools to create a custom question answering project and allows you to upload knowledge base documents like FAQs and manuals.
- B. Incorrect.
Incorrect: Training a custom language model with Azure Machine Learning is unnecessary for the question answering solution. Azure AI Language provides pre-built capabilities for question answering.
- C. Correct.
Correct: Once the custom question answering project is created, it must be published to an endpoint so it can be used by client applications.
- D. Correct.
Correct: Testing the deployed endpoint with sample queries ensures that the solution functions as expected and delivers accurate responses.
- E. Incorrect.
Incorrect: While Azure AI Language supports multilingual features, enabling the Translation feature is not directly related to creating a custom question answering solution. Multilingual support can be configured separately if needed.