Vector Embeddings and Semantic Search Fundamentals

1. True or False: Two semantically similar documents will have embeddings with high cosine similarity.

True

False

2. True or False: Semantic search using embeddings can understand synonyms and related concepts without explicit keyword matching.

True

False

3. Which of the following is a limitation of vector embeddings?

They cannot represent numerical data

They require significant computational resources to generate and store

They only work with small datasets

They are incompatible with modern hardware

4. How does semantic search improve the user experience in search applications?

By returning results matching user intent, not just keywords

By displaying results in reverse alphabetical order

By limiting results to single-word queries

By removing all filtering options

5. What is the typical range of dimensions used in modern text embeddings?

1–10 dimensions

50–500 dimensions

1,000–100,000 dimensions

Only 1 dimension is standard

6. True or False: Fine-tuning a pre-trained embedding model can improve performance on domain-specific tasks.

True

False

7. In semantic search, the query is converted to an embedding and compared against which of the following?

Raw text documents only

Embeddings of documents in a vector database

SQL database schemas

Image pixel values

8. What is the primary advantage of using vector embeddings over bag-of-words representations?

Embeddings capture semantic context and relationships

Embeddings require less computational power

Embeddings eliminate the need for preprocessing

Embeddings work only with English text

9. Which metric is most commonly used to measure distance between embeddings in semantic search?

Manhattan distance

Euclidean distance

Cosine similarity

Hamming distance

10. True or False: Vector embeddings can only represent text data, not images or audio.

True

False

11. What is a vector embedding in the context of machine learning?

A numerical representation of text or objects in a high-dimensional space

A physical storage device for storing large datasets

A programming framework for building neural networks

A method for compressing image files

12. How do pre-trained embeddings benefit semantic search applications?

They eliminate the need for training data

They provide ready-made representations capturing semantic relationships

They guarantee 100% accuracy in all queries

They reduce storage costs to zero

13. Transformer models like BERT and GPT generate embeddings by using which mechanism?

Attention mechanisms

Hash tables

Binary search trees

Sorting algorithms

14. What is a vector database used for in semantic search applications?

Storing and querying embeddings efficiently

Replacing traditional SQL databases

Managing user passwords

Compressing video files

15. Which of the following best describes dimensionality in vector embeddings?

The number of features or components in each embedding vector

The physical size of the storage device

The number of users accessing the system

The speed of network connections

16. In semantic search, what role do embeddings play?

They encrypt sensitive data

They convert queries and documents into comparable numerical forms

They eliminate the need for databases

They reduce query execution time by 100%

17. Embeddings are typically generated by training which type of model?

Decision trees

Neural networks or deep learning models

Linear regression

Rule-based expert systems

18. What does cosine similarity measure in vector embeddings?

The distance between two points in Euclidean space

The angle between two vectors and their directional similarity

The total magnitude of a single vector

The compression ratio of embedding data

19. Which of the following is a common use case for vector embeddings?

Building recommendation systems

Storing relational database schemas

Compiling source code

Managing file permissions

20. How does semantic search differ from keyword-based search?

It searches for exact word matches only

It understands the meaning and intent behind queries

It requires more storage space than keyword search

It only works with numerical data

DataAI Vector Embeddings and Semantic Search Quiz

1. True or False: Two semantically similar documents will have embeddings with high cosine similarity.

2.

What first name or nickname would you like us to use?

2. True or False: Semantic search using embeddings can understand synonyms and related concepts without explicit keyword matching.

3. Which of the following is a limitation of vector embeddings?

4. How does semantic search improve the user experience in search applications?

5. What is the typical range of dimensions used in modern text embeddings?

6. True or False: Fine-tuning a pre-trained embedding model can improve performance on domain-specific tasks.

7. In semantic search, the query is converted to an embedding and compared against which of the following?

8. What is the primary advantage of using vector embeddings over bag-of-words representations?

9. Which metric is most commonly used to measure distance between embeddings in semantic search?

10. True or False: Vector embeddings can only represent text data, not images or audio.

11. What is a vector embedding in the context of machine learning?

12. How do pre-trained embeddings benefit semantic search applications?

13. Transformer models like BERT and GPT generate embeddings by using which mechanism?

14. What is a vector database used for in semantic search applications?

15. Which of the following best describes dimensionality in vector embeddings?

16. In semantic search, what role do embeddings play?

17. Embeddings are typically generated by training which type of model?

18. What does cosine similarity measure in vector embeddings?

19. Which of the following is a common use case for vector embeddings?

20. How does semantic search differ from keyword-based search?