ADA-C01 Question 191
Single answerImplement data classification (EXTRACT_SEMANTIC_CATEGORIES, ASSOCIATE_SEMANTIC_CATEGORIES)A Snowflake administrator needs to classify a newly ingested CUSTOMER_PROFILE table before security teams apply masking policies. The team wants to use Snowflake's semantic category workflow so they can first review the detected classifications and then persist approved associations to the table columns. Which approach best meets this requirement?
- A
Run EXTRACT_SEMANTIC_CATEGORIES on the CUSTOMER_PROFILE table to detect likely semantic categories, review the returned results, and then use ASSOCIATE_SEMANTIC_CATEGORIES to write the approved category associations to the relevant columns.
- B
Run ASSOCIATE_SEMANTIC_CATEGORIES on the CUSTOMER_PROFILE table first to detect candidate semantic categories, and then run EXTRACT_SEMANTIC_CATEGORIES only for columns that were not classified.
- C
Use EXTRACT_SEMANTIC_CATEGORIES to permanently tag the CUSTOMER_PROFILE columns, because ASSOCIATE_SEMANTIC_CATEGORIES is only used to export classification results for audit review.
- D
Use ALTER TABLE ... SET TAG statements instead of semantic category functions, because Snowflake semantic categories are generated automatically during data load and cannot be reviewed before they are applied.
Show answer and explanation
Correct answer: A
Explanation
The best answer is to use a two-step workflow: first detect, then associate. In Snowflake's data classification process, EXTRACT_SEMANTIC_CATEGORIES is used to inspect table data and return candidate semantic categories for columns. This allows governance or security teams to review the findings before making changes. ASSOCIATE_SEMANTIC_CATEGORIES is then used to persist approved semantic category associations to the columns. This approach is aligned with best practice because it separates discovery from application, reducing the risk of incorrectly classifying sensitive data. Candidates should recognize that EXTRACT_SEMANTIC_CATEGORIES is not itself the persistence mechanism, and that semantic category association is not automatic on load. Refer to Snowflake documentation for data classification and semantic category association workflows, including the usage of EXTRACT_SEMANTIC_CATEGORIES and ASSOCIATE_SEMANTIC_CATEGORIES in governance-driven review processes.
- A. Correct.
Correct. EXTRACT_SEMANTIC_CATEGORIES is used to analyze data and return inferred semantic categories without persisting them to the columns. This supports a review-and-approve workflow. After validation, ASSOCIATE_SEMANTIC_CATEGORIES can be used to associate the approved semantic categories with the appropriate columns. This is the practical sequence when teams want detection first and controlled application second.
- B. Incorrect.
Incorrect. This reverses the purpose of the functions. ASSOCIATE_SEMANTIC_CATEGORIES does not perform initial discovery by itself in the way described; it is used to associate semantic categories, typically after review of extracted results. EXTRACT_SEMANTIC_CATEGORIES is the discovery step, not the cleanup step for unclassified columns.
- C. Incorrect.
Incorrect. EXTRACT_SEMANTIC_CATEGORIES does not permanently apply classifications to columns. It returns inferred semantic category information for inspection. The misconception here is treating extraction as a write operation, when it is actually intended for analysis and review.
- D. Incorrect.
Incorrect. Standard object tags and semantic categories are not the same workflow. While tags are used broadly in governance, the semantic category process specifically uses Snowflake-provided classification functions. Also, semantic categories are not automatically generated and applied simply because data was loaded. Administrators can review extracted results before associating them.