AI-102 Question 296
Select 4You are tasked with creating a custom question answering project for an internal knowledge base using Azure AI services. The knowledge base contains documents in multiple formats, including PDFs and Word files. You need to design the solution so it enables users to query the knowledge base efficiently. Which steps must you perform to successfully create and deploy the custom question answering project?
- A
Upload the documents to Azure Blob Storage and configure Cognitive Search to index the content.
- B
Create a new Language Studio project and select 'Question Answering' as the project type.
- C
Define question-answer pairs manually in Language Studio if no documents are uploaded.
- D
Train the model using Azure Machine Learning before deploying the custom question answering project.
- E
Publish the knowledge base to make it accessible via an endpoint.
Show answer and explanation
Correct answers: A, B, C, E
Explanation
To create a custom question answering project, you must first upload the documents or define question-answer pairs manually in Language Studio. Then, configure the project in Language Studio, which includes indexing content (with Cognitive Search if needed). Finally, publish the project to make it accessible. Azure Machine Learning is not required for this process because Azure Language Service handles the training and deployment automatically.
- A. Correct.
Correct. Uploading the documents to Azure Blob Storage and configuring Cognitive Search allows the system to process and index the unstructured content into a searchable format.
- B. Correct.
Correct. Creating a project in Language Studio and selecting 'Question Answering' is the necessary first step to configure the custom question answering project.
- C. Correct.
Correct. If no documents are uploaded, you can manually define question-answer pairs within Language Studio to populate the knowledge base.
- D. Incorrect.
Incorrect. Azure Machine Learning is not required for training a custom question answering project since the Language Service manages the training automatically.
- E. Correct.
Correct. Publishing the knowledge base is required to expose the functionality via an endpoint so users can query it.