NEW
Font size
WorksheetsNatural Language Processing Quiz
Total questions: 20
Worksheet time: 10mins
What is the main goal of Natural Language Processing (NLP)?
Translate human language into computer code
Create new programming languages
Bridge the gap between human communication and machine understanding
Improve image recognition accuracy
Which of the following is NOT a core area of NLP?
Speech recognition
Machine translation
Image segmentation
Text processing
Why is text preprocessing important in NLP?
It removes valuable insights
It makes the text harder to interpret
It reduces data quality
It cleans messy data and improves model accuracy
What does tokenization do?
Encrypts the text
Splits text into smaller units like words or sentences
Translates text into another language
Summarizes long texts
Which type of tokenization is used by models like BERT or GPT?
Sentence tokenization
Word tokenization
Subword tokenization
Character tokenization
What is the risk of removing all stopwords during preprocessing?
It makes training faster
It increases vocabulary size
It can remove important words like 'not' and change meaning
It improves context understanding
What is Part-of-Speech (POS) tagging?
Removing unwanted characters from text
Splitting text into subwords
Assigning grammatical categories to each word
Detecting spam messages
What does Named Entity Recognition (NER) identify in text?
Spelling errors
Sentiment polarity
Named entities like people, dates, and locations
Text length
What is the purpose of vectorization in NLP?
Translate text into speech
Convert text into numerical format for machine learning
Compress text files
Detect named entities
Which method considers the importance of words across the entire corpus?
Bag of Words
Lemmatization
Word Embeddings
TF-IDF
Why are word embeddings preferred over BoW?
They reduce training time
They are easier to compute
They capture semantic meaning and word relationships
They do not require any preprocessing
Which traditional ML model is commonly used for binary text classification?
Decision Trees
Logistic Regression
Naive Bayes
SVM
What makes Naive Bayes 'naive'?
It ignores punctuation
It assumes words are dependent
It uses no training data
It assumes word occurrences are independent
What is a key advantage of Naive Bayes?
Requires large datasets
Performs poorly on text data
Fast and effective on small datasets
Always requires embeddings
What is the main idea behind Support Vector Machines (SVM)?
Predict sequences with high accuracy
Find the optimal decision boundary that separates classes
Randomly classify data
Use Naive assumptions about text
What are support vectors in SVM?
Vectors that contain stopwords
The points farthest from the hyperplane
The closest data points to the decision boundary
Words that represent vector embeddings
What makes Random Forest more robust than a single decision tree?
It always gives the same output
It uses multiple trees and combines their outputs
It removes the need for preprocessing
It trains slower than trees
What is a major limitation of basic RNNs?
Too fast for training
Limited accuracy on structured data
Vanishing gradient problem with long sequences
Inability to work with text
What does the Forget Gate in LSTM do?
Decides what information to discard from memory
Stores the input
Prevents dropout
Tracks activation functions
What is the main difference between GRU and LSTM?
GRU has fewer gates and is computationally simpler
LSTM is used only for images
GRU requires labeled data
LSTM cannot handle sequences
