Building LLM Applications with RAG and Vector Search →hardMultiple ChoiceObjective-mapped

1Z0-1127-25 Practice Question: Building LLM Applications with RAG and Vector Search

Exhibit

Refer to the exhibit.

```sql
-- Oracle Database 23ai AI Vector Search index creation
CREATE VECTOR INDEX doc_vec_idx ON documents(chunk_embedding) 
  ORGANIZATION NEIGHBOR PARTITIONS
  DISTANCE COSINE
  WITH TARGET ACCURACY 95
  PARAMETERS (TYPE IVF, NEIGHBOR PARTITIONS 4);
```

A DBA has created the above vector index. After running queries, they observe that recall is lower than expected for approximate searches. Which change would most likely improve recall while maintaining query performance?

⚠ Common exam trap

Oracle often tests the misconception that changing the index type (e.g., IVF to HNSW) is the primary way to fix recall, when in fact TARGET ACCURACY is the direct parameter for recall tuning without altering the index structure.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Increase the TARGET ACCURACY value to 99.

Increasing TARGET ACCURACY to 99 directly raises the recall threshold for approximate search, forcing the vector index to retrieve more candidates during the search phase. This improves recall without changing the index structure or query parallelism, so query performance (latency) is only minimally impacted compared to switching index types or drastically altering neighbor partitions.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

✗
Change the index type from IVF to HNSW.
Why it's wrong here
While HNSW generally offers better recall, the question asks for a change within the current index context; TARGET ACCURACY is more direct.
✓
Increase the TARGET ACCURACY value to 99.
Why this is correct
A higher TARGET ACCURACY forces the approximate search to consider more vectors, increasing recall at the cost of some latency.
✗
Increase the number of neighbor partitions (NEIGHBOR PARTITIONS) to 8.
Why it's wrong here
More partitions can improve parallelism but may not directly improve recall; recall is controlled by TARGET ACCURACY.
✗
Reduce the number of neighbor partitions to 2.
Why it's wrong here
Fewer partitions reduce the search space, likely lowering recall.

Go deeper

1Z0-1127-25 Practice Question: Building LLM Applications with RAG and Vector Search

Exhibit

Why each option matters

Option-by-option breakdown

Related to this question