1Z0-184-25 Question 166
Select 2You have an Autonomous Database configured to allow natural language queries using Select AI. Your marketing department wants to run ad-hoc queries like: 'Which product categories had the highest revenue last quarter?' against a table named SALES_DATA with columns such as PRODUCT_CATEGORY, REVENUE, and QUARTER. Which two steps must you take so these natural language queries work correctly in the Select AI interface?
- A
Grant the SELECT_CATALOG_ROLE to the marketing team so they can access the data dictionary metadata.
- B
Enable the �Natural Language Querying� feature for the database instance in the Autonomous Database console under Select AI settings.
- C
Create or verify metadata descriptions or synonyms for key tables and columns so the AI engine can interpret table references correctly.
- D
Flatten your SALES_DATA table columns into a single CLOB field to give the AI engine unstructured text for processing.
Show answer and explanation
Correct answers: B, C
Explanation
To make Select AI work effectively with natural language, you must enable the feature in the Autonomous Database console and ensure that your table and column references are clearly defined for the AI engine. Accurate metadata descriptions or synonyms improve query interpretation. Granting the SELECT_CATALOG_ROLE alone does not accomplish this, nor does converting properly structured data into a single unstructured field. For further guidance, see Oracle� Autonomous Database and Select AI documentation on configuring metadata and enabling natural language features.
- A. Incorrect.
Option 1: Incorrect. SELECT_CATALOG_ROLE alone does not enable natural language querying in Select AI. While controlling dictionary access might be important for metadata insights, simply granting this role is insufficient to allow AI-based natural language queries.
- B. Correct.
Option 2: Correct. You must enable the �Natural Language Querying� or equivalent AI features in the Autonomous Database console to allow the usage of Select AI. This activates the underlying feature that interprets plain-text prompts.
- C. Correct.
Option 3: Correct. Providing accurate metadata or synonyms helps the AI engine map natural language terms (like 'product category') to the actual table and column names, improving query accuracy.
- D. Incorrect.
Option 4: Incorrect. Flattening data into an unstructured CLOB typically reduces the AI engine� ability to leverage relational structure and column definitions. Select AI can natively understand the structure if it is properly configured through metadata.