What is the primary goal of Natural Language Processing?

• To make computers faster
• To enable machines to understand and process human language
• To build better hardware
• To create computer viruses

The process of breaking a sentence into individual words is called:

• Parsing
• Tokenization
• Stemming
• Filtering

Which technique removes common words like 'is', 'the', and 'a'?

• Lemmatization
• Stop Word Removal
• POS Tagging
• NER

Reducing 'running' to 'run' using a crude heuristic is known as:

• Stemming
• Lemmatization
• Tokenization
• Vectorization

What is a collection of text documents used for analysis called?

• Database
• Corpus
• Dictionary
• Frame

What is the main advantage of Word2Vec over Bag of Words?

• It is faster to train
• It captures semantic relationships and meaning
• It uses less memory
• It only works on English

In Word2Vec, the 'Skip-gram' model predicts:

• Target word from context
• Context words from a target word
• The next sentence
• The sentiment

Which similarity measure is most effective for comparing word vectors?

• Euclidean Distance
• Cosine Similarity
• Jaccard Index
• Manhattan Distance

Word Sense Disambiguation' is used to:

• Find typos
• Determine the correct meaning of a word in context
• Count syllables
• Translate words

What does 'BLEU' score evaluate?

• Image quality
• Machine Translation quality
• CPU performance
• Database speed

Which mechanism allows a Transformer to focus on different parts of a sentence?

• Backpropagation
• Self-Attention
• Recursion
• Convolution

What is the core architecture in the 'Attention is All You Need' paper?

• RNN
• CNN
• Transformer
• LSTM

BERT is an acronym for:

• Binary Encoder Representation
• Bidirectional Encoder Representations from Transformers
• Basic Entity Recognition Tool
• Binary Entity Relation Task

What is 'Masked Language Modeling' (MLM)?

• Hiding user identities
• Predicting words that are hidden in a sentence
• Translating hidden text
• Removing noise

In a Transformer, 'Positional Encoding' is needed because:

• It stores the file path
• The model has no inherent sense of word order
• It encrypts data
• It compresses the model