1Z0-184-25 Question 33
Single answerYou are building a product recommendation engine on Oracle Cloud Infrastructure. Your table stores text embeddings in a vector column, and you want to enable efficient vector-based similarity queries. Which action should you take to define an appropriate index on the vector column?
- A
Drop and recreate your table to specify a vector index during table creation, ensuring the vector column is indexed appropriately.
- B
Use a standard B-tree index on the vector column using the standard CREATE INDEX syntax for numeric data.
- C
Leverage a specialized index specifically for vectors by referencing the vector� dimension and distance measure, for example, using CREATE VECTOR INDEX on the vector column.
- D
Set a hidden database parameter to allow the existing numeric index to support approximate vector similarity queries.
Show answer and explanation
Correct answer: C
Explanation
To efficiently perform vector similarity searches, Oracle recommends using a specialized vector index that understands the dimension and distance metric of the vector data. This approach is documented in Oracle� Database and OCI documentation for vector data types and indexing, enabling fast, accurate retrieval of items based on vector similarity.
- A. Incorrect.
Option 1 is incorrect because dropping and recreating the entire table is unnecessary. Oracle allows you to create or add a specialized index on an existing vector column without fully recreating the table.
- B. Incorrect.
Option 2 is incorrect because B-tree indexes are not well suited for vector similarity searches. These indexes optimize equality or range queries, not multi-dimensional similarity queries.
- C. Correct.
Option 3 is correct. For efficient vector similarity queries, you must use a specialized vector index that includes specifying the vector� dimension and the appropriate distance measure. This ensures the query optimizer can perform similarity operations efficiently on the vector data.
- D. Incorrect.
Option 4 is incorrect. Vector similarity queries require specialized indexing techniques, not a hidden parameter toggling or a reliance on numeric indexes. Simply setting a parameter does not convert a standard index to handle vector data appropriately.