AI-102 Question 295
Single answerYou are tasked with creating a custom question answering project using Azure AI Language service for your organization. The organization wants to use existing knowledge bases, such as manuals and FAQ documents, to answer customer queries. What is the correct process to configure and publish the custom question answering project?
- A
Upload documents to Azure Blob Storage and directly query them using the Azure Question Answering REST API.
- B
Use the Azure Language Studio to create a new Question Answering project, upload the documents, configure the project, and then deploy the project.
- C
Manually extract questions and answers from the documents and upload them as a JSON file to Azure Cognitive Search.
- D
Train a custom machine learning model using Azure Machine Learning to process the documents and create a question answering capability.
Show answer and explanation
Correct answer: B
Explanation
The correct process for creating a custom question answering project involves using Azure Language Studio, where you can upload documents (such as manuals and FAQs), configure the knowledge base, and deploy it to handle customer queries. Azure Language Studio simplifies the process by leveraging pre-built AI capabilities for question answering, eliminating the need for manual processing or custom model training.
- A. Incorrect.
Uploading documents to Azure Blob Storage does not automatically enable question answering capabilities. You need to use Azure Language Studio to create and configure the custom question answering project.
- B. Correct.
This is the correct process. Azure Language Studio allows you to create a custom question answering project by uploading knowledge base documents, configuring the project, and deploying it for use via the REST API or client applications.
- C. Incorrect.
Manually extracting questions and answers is not required. Azure's Question Answering service can automatically process and extract relevant information from uploaded documents.
- D. Incorrect.
While Azure Machine Learning can be used for custom AI models, it is not the appropriate approach for creating a custom question answering project. Azure Language service provides a more streamlined and specialized solution for this purpose.