NEW
Font size
WorksheetsFeature Extraction Quiz (BoW & TF-IDF)
Total questions: 20
Worksheet time: 10mins
Which of the following best describes the Bag of Words (BoW) model?
It captures the order of words in a sentence.
It uses semantic relationships between words.
It converts text into a vector based on word frequency.
It reduces the dimensionality of feature vectors using PCA.
Which technique assigns lower weights to common words like 'the' or 'is'?
BoW
TF
TF-IDF
One-hot encoding
What does TF in TF-IDF stand for?
Term Frequency
Textual Frequency
Total Frequency
Token Frequency
What is the purpose of IDF in TF-IDF?
To normalize word frequencies
To highlight frequent terms in the corpus
To penalize terms that appear in many documents
To tokenize the input text
Which of the following is a limitation of BoW?
It cannot be used for long documents.
It is not applicable to English language.
It ignores word order and context.
It works only with numerical data.
Which method could address the limitation of BoW by capturing phrases?
Stemming
N-grams
Stopword removal
Lemmatization
Which of the following has the highest IDF value?
A word that appears in every document
A word that appears in no documents
A word that appears in only one document
A word that appears in half the documents
TF-IDF is generally preferred over BoW because:
It uses deep learning.
It requires no preprocessing.
It balances term frequency with document rarity.
It captures the meaning of sentences.
What does a TF-IDF score of zero mean?
The term is very rare.
The term does not appear in the document.
The term appears in all documents.
The term has no semantic meaning.
Which component of TF-IDF varies depending on the document itself?
Term Frequency
Inverse Document Frequency
Corpus size
Document length
Which technique represents each word with a binary 0 or 1?
TF-IDF
BoW
Word2Vec
One-hot encoding
Which term in TF-IDF penalizes common words?
TF
IDF
N-gram
Token
What is the main mathematical operation in computing IDF?
Multiplication
Subtraction
Logarithm
Division
BoW creates a vector whose length equals to:
Number of words in the document
Number of documents in the corpus
Total unique words in the corpus
Total number of characters in the document
Which of the following is most impacted by document length?
TF
IDF
TF-IDF
BoW
In TF-IDF, the log base in IDF is usually:
2
10
e
Any positive base
Why might you use stopword removal before applying BoW?
To reduce document length
To remove punctuation
To eliminate common, less informative words
To tokenize the text
What is a major disadvantage of BoW in sentiment analysis?
Cannot be used in English
Fails to distinguish between opposite meanings like 'good' and 'not good'
Requires labeled data
Depends heavily on word embeddings
Which step is NOT part of creating a BoW model?
Tokenization
Vectorization
Synonym replacement
Vocabulary building
What happens to rare words in a TF-IDF representation?
They are removed
They get lower weights
They get higher weights
They are ignored
