Exam 1Z0-184-25: Oracle AI Vector Search Professional

1. What is the significance of splitting text into chunks in the process of loading data into Oracle Al Vector Search?

To reduce the computational burden on the embedding model

To facilitate parallel processing of the data during vectorization

To minimize token truncation as each vector embedding model has its own maximum token limit<<<

2. Which operation is NOT permitted on tables containing vEcroR columns?
SELECT
UPDATE
DELETE
JOIN on vEcToR columns <<<

3. What happens if a vecmop column contains vectors of different dimensions and an attempt is made to create a vector index on it?
The index creation succeeds, but with reduced accuracy.
The database automatically normalizes the vectors
An error is thrown, preventingindex creation. <<<
The vectors with higher dimensions are truncated,

4. What is created to facilitate the use of 0Cl Generative Al with Autonomous Database?
An Al profile for OCl Generative Al <<<
A dedicated OCl compartment
A new user account with elevated privileges

5. What is a key benefit of Select Al?
Increased complexity in querying data
The ability to use natural language to query data <<<
A requirement to learn complex SOL queries

6. What signifies a new era of Al capabilities in Exadata System Software 24ai?
Automated data labeling for machine learning tasks

Integration with open-source Al frameworks
Key features like Al Smart Scan and infrastructure improvements <<<

7. What does the FErcH APPRoxIMArE clause accomplish in a similarity search?
Guarantees exact matches with faster performance
FETCH APPRoxIMarE enables the query optimizer to use a vector index <<<

Ensures only the most accurate results are returned

Filters results by their similarity score thresholds

8. What is the primary function of Al Smart Scan in Exadata System Software 24ai?
To provide real-time monitoring and diagnostics for Al applications

 To accelerate Al workloads by leveraging Exadata RDMA Memory (XRMEM), Exadata Smart Flash Cache, and on-storage processing <<<

To automatically optimize database queries for improved performance

9. Which of the following SOL queries uses Euclidean Squared Distance for a similarity search?
SELECT
FROM vector_tab
ORDER BY VECTOR DISTANcE(embedding,:query vector,COSINE)

FETCH FIRST 10 ROS ONY

SELECT
FROM vector tab
ORDER BY VECTOR DISTANCE(embedding,:query vector,EUCLIDEAN) 

FETCH FIRST 10 ROMS ONLY;

SELECT
FROM vectortak
ORDER BY VECTOR_DISTANCE(embedding,:guery vector,EUCLIDEAN_SURED) <<<

FETCH FIRST 10 ROWS ONLY:

SELECT docID
FROM vetor_tab
WHERE VECTOR DISTANCE(embedding,:query vector,EUCLIDEAN)<10

10. What is a primary difference between HNsw and ive vector indexes in Oracle Database 23ai?
HNsw is used for disk-based searches, whereas Ive is memory-based.
HNsw supports hierarchical graphs, whereas Ivr uses partition-based clustering. <<<
HNsw relies on Euclidean distance only, whereas Ive supports cosine similarity only.
HNsw is used for exact searches, whereas Ivr is used for approximate searches.

11. What security enhancement is introduced in Exadata System Software 24ai?
Integration with third-party security tools
Enhanced encryption algorithms for data at rest
SNMP Security<<<

12. Which type of vector index is best suited for larger datasets that might not fit in memory?
O HNSW
O IVF  <<<
O Both HNSW and IVF are equally suitable for large datasets

13. Which feature enhances RDMA over converged Ethernet (RoCE) network resilience in Exadata System Software 24ai?
OAWR and SOL Monitor Enhancements
OEnhanced RoCE Network Discovery
O lmproved RoCE Network Resilience <<<

14. You want to quickly retrieve the top-10 matches for a query vector from a dataset of bilions of vectors, prioritizing speed over exact accuracy
What is the best approach?
Exact similarity search using flat search
Approximate similarity search with a low target accuracy setting <<<
Relational filtering combined with an exact search
Exact similarity search with a high target accuracy setting

15. Which of the following is a valid dimension format for a vector data type in Oracle Database 23ai?
OFLOAT32 <<<
O STRING
O BOOLEAN

16. What is a primary advantage of using approximate similarity search?
Guaranteed accuracy of search results

Faster search times in large datasets <<<
Simplified SQL syntax for vector searches

Support for all distance metrics equally

17. Which initialization parameter must be configured to enable the creation of vector indexes in Oracle Database 23ai?
SGA
VECTOR MEMORY SIZE<<<
VECTOR
QUERY PARALLELISM

18. Which SOL statement correctly adds a vecrop column named v with 3 dimensions and uoar32 format to an existing table named my table?
OALTER TABLE mY table ADD(v VECTOR(3,FLOAT32))
OALTER TABLE mY table MODIFY(V VECTOR(3,FLOAT32))
ALTER TABLE my_table ADD V VECTOR(3,FLOAT32) <<<
OUPDATE mY table SET v=VECTOR(3,FLOAT32)

19. A data scientist is creating a vector index on a dataset of 1 milion vectors, Thev observe that the indexing process is slow
Which parameter adjustment could reduce indexing time for an lVF index without significantly compromising accuracy?
Increasing the VECTOR MEMORY SIZE
Decreasing the number Of NEIGHBOR PARTITIONS <<<
Setting alower TARGET ACCURACY
Using the Dor distance metric instead of cosINE

20. What is the advantage of using local ONNX models for embedding within the database?
Enhanced security, as no data is sent outside the database <<<
Increased flexibility in choosing embedding models from various sources

Simplified integration with external machine learning libraries
lmproved performance by leveraging cloud-based computational resources

21. What is the purpose of a vector pool in Oracle Database 23ai?
To store all database data
To store HNSW vector indexes and associated metadata <<<

To store 3rd party embedding models

22. In RAG, what is the role of the vector database?
◎To store private content which can be used to enhance a user's query <<<

To provide general knowledge to the LLM.
To generate text responses directly to the user.

23. A developer is running a query with an aNsw index and notices that the search is returning fewer results than expected.
What could be the cause?
The VECTOR MEMORY SIZE is too large
The EFSEARCH parameter value is too low. <<<
The query is using an incompatible distance metric.
The index was created with NEIGHBoRs set to the maximum yalue

24,Which statement accurately describes the function of the dbms vectorchain.utl to embeddings()function in Oracle Database 23ai?
lt facilitates the compression of large text datasets for efficient storage.
It encodes unstructured data in vectors. <<<
◎ lt generates synthetic data to augment training datasets for Al models.

25. What is the key purpose of a multi-vector similarity search?

Performing searches that prioritize accuracy over speed

Searching across multiple vector datasets simultaneously

Finding top-k matches grouped by partitions such as documents or categories <<<

Enhancing relational filters with vector search capabilities

26. What happens when you atempt to insert a vector with an incorect number of dimensions into a vEcroR column with a defined number of dimensions?
OThe database truncates the vector to fit the defined dimensions.
O The database pads the vector with zeros to match the defined dimensions.、
OThe insert operation fails, and an error message is thrown.<<<
OThe database ignores the defined dimensions and inserts the vector as is

27. Which distance metric measures the angle between two vectors?
Euclidean
OManhattan
O Cosine <<<

28. What is the primary purpose of the Python code snippets provided in the context of Oracle Vector Search and 0Cl Generative A!?
To demonstrate how to create a custom embedding model from scratch
O) To showcase how to leverage Oracle Al Vector Search to retrieve contextually relevant information for augmenting LLM prompts  <<<

O To explain the steps involved in setting up and configuring an Oracle Database instance for Al workloads

29. What is a primary use case for GoldenGate in the context of Al?
To create static backups of data

To prevent data from being used by Al applications.

To replicate vector changes across multiple databases

30. What is the primary purpose of using Retrieval Augmented Generation (RAG) with Al Vector Search in Oracle Database 23ai?
To train the LLM with new data from Oracle Database 23ai
To provide the LLM with relevant context from Oracle Database 23ai, enabling it to generate more accurate responses  <<<

To store the entire LLM model within Oracle Database 23ai for faster processing
To replace traditional keyword-based search with LLM-powered semantic search in Oracle Database 23ai