NLP Concepts Quiz

What is the main limitation of one-hot encoding in NLP?

• It is too short

• It captures semantics well

• It cannot represent numerical data

• It does not capture semantic relationships

Which of the following is a frequency-based method for word representation?

• Word2Vec

• BERT

• Count Vector (Bag of Words)

• GPT

Word embeddings map words to:

• Integers

• Binary values

• High-dimensional real-valued vectors

• One-hot positions

Which technique improves semantic similarity representation?

• One-hot

• Hashing

• Word Embedding

• Symbolic logic

Which is NOT a word embedding model?

• Word2Vec

• GloVe

• TF-IDF

• BERT

Why are word embeddings important in NLP?

• They are easy to visualize

• They reduce memory

• They help models capture context and meaning

• They simplify algorithms

A neural network is inspired by:

• Animal instincts

• Artificial neurons

• Human reflex

• Brain's structure and functioning

What adjusts during the training of a neural network?

• Bias

• Threshold

• Weights

• Inputs

In a neural network, an edge connects:

• Layers and activations

• Inputs and outputs

• Neurons

• Datasets

CNNs are commonly used in NLP for:

• Machine translation

• Sentiment analysis and text classification

• Parsing

• Question answering