AI-102 Question 385
Select 3You are designing an Azure AI solution that includes a Knowledge Store to store enriched data from an Azure Cognitive Search index. You need to configure projections to enable different types of data to be exported for further analysis. Which of the following steps must you take to manage the projections correctly?
- A
Define the desired output structure in the Knowledge Store JSON configuration file.
- B
Enable the 'Skillset' option in Azure Cognitive Search to automatically create projections.
- C
Specify the file, object, or table projection type in the Knowledge Store definition.
- D
Set up a connection to an Azure Data Lake Storage account for storing projections.
- E
Manually trigger the enrichment pipeline to generate projections.
Show answer and explanation
Correct answers: A, C, D
Explanation
To manage Knowledge Store projections, you must define the output structure within the JSON configuration file, specify the type of projection (file, object, or table), and ensure a connection to an Azure Data Lake Storage account is set up to store the exported data. These steps are critical for properly managing and utilizing Knowledge Store projections in Azure Cognitive Search.
- A. Correct.
Defining the desired output structure in the Knowledge Store JSON configuration file is essential to specify how the data will be exported and stored. This step is required for managing projections.
- B. Incorrect.
While a skillset is necessary for data enrichment, enabling it does not automatically create projections. Projections need to be explicitly defined in the Knowledge Store configuration.
- C. Correct.
Specifying the projection type (file, object, or table) in the Knowledge Store definition is required to determine the format and structure of the exported data.
- D. Correct.
A connection to an Azure Data Lake Storage account is necessary because the Knowledge Store uses it to store the projection output.
- E. Incorrect.
Manually triggering the enrichment pipeline is not required for managing projections. The pipeline runs automatically when indexing is configured and executed.