AI-102 Question 386
Select 3You are designing a Knowledge Store for a search solution using Azure Cognitive Search. The content being indexed includes both structured and unstructured data. You need to create projections to store the enriched data for downstream analytics and reporting. Which combination of projections should you use to manage the data effectively?
- A
File projections to store the unstructured content like PDFs or images.
- B
Object projections to organize the enriched data into a hierarchical JSON format.
- C
Table projections to enable structured querying and reporting in relational formats.
- D
File projections to transform structured data into a relational format.
- E
Object projections to convert relational data into binary files for storage.
Show answer and explanation
Correct answers: A, B, C
Explanation
When managing Knowledge Store projections, each projection type serves a specific purpose: file projections handle unstructured content like PDFs and images, object projections organize data in a hierarchical JSON format, and table projections enable structured querying in tabular formats. Using the correct combination of projections ensures the enriched data can be effectively used for different downstream tasks.
- A. Correct.
Correct. File projections are suitable for storing unstructured content, such as PDFs or images, that need to be preserved in their original format for downstream processing or archival purposes.
- B. Correct.
Correct. Object projections are ideal for organizing enriched data into hierarchical JSON structures, which can be useful for applications that require JSON-formatted outputs.
- C. Correct.
Correct. Table projections are used to transform enriched data into a structured, tabular format, making it easy to query using relational database tools.
- D. Incorrect.
Incorrect. File projections are not used to transform structured data into relational formats. This is the role of table projections.
- E. Incorrect.
Incorrect. Object projections are not designed to convert relational data into binary files. They are used for organizing data into JSON-based hierarchical structures.