Sentiment analysis basics quiz

Question 1

What is sentiment analysis?

Accepted Answer

Detecting the feeling or opinion in text

Answer

Counting how many words are in a sentence

Answer

Translating text into another language

Answer

Checking spelling mistakes

Question 2

The three most common categories used in basic sentiment analysis are positive, negative, and what third one?

Accepted Answer

Neutral

Answer

Loud

Answer

Fast

Answer

Colorful

Question 3

Which sentence would most likely be labeled as POSITIVE sentiment?

Accepted Answer

I absolutely love this movie, it was amazing!

Answer

This is the worst meal I have ever eaten.

Answer

The bus arrives at 9 in the morning.

Answer

I will never shop here again.

Question 4

How does a computer 'learn' to do sentiment analysis from examples?

Accepted Answer

By being shown many labeled example texts and finding patterns

Answer

By memorizing the dictionary

Answer

By asking a human every single time

Answer

By randomly guessing without any data

Question 5

In machine learning for sentiment analysis, what is 'labeled data'?

Accepted Answer

Text that already has the correct sentiment tag attached

Answer

Text written in capital letters

Answer

Data stored on a label printer

Answer

Any text without punctuation

Question 6

What is the 'training data' used for in a sentiment analysis model?

Accepted Answer

The labeled examples the model studies to learn patterns

Answer

The final answer the model prints

Answer

The computer's electricity supply

Answer

The user typing on a keyboard

Question 7

What does 'tokenization' mean in text processing for sentiment analysis?

Accepted Answer

Breaking a sentence into smaller pieces like words

Answer

Adding more words to a sentence

Answer

Encrypting text with a password

Answer

Deleting all the words

Question 8

A simple 'word list' (lexicon) approach scores sentiment by doing what?

Accepted Answer

Counting known positive and negative words in the text

Answer

Measuring the length of the text

Answer

Checking the font color of the text

Answer

Counting the number of sentences only

Question 9

Why is the sentence 'The movie was not good' tricky for a simple word-counting model?

Accepted Answer

The word 'not' flips the meaning of the positive word 'good'

Answer

The sentence is too long to read

Answer

There are no real words in it

Answer

It contains a spelling mistake

Question 10

Why is sarcasm a hard problem for sentiment analysis?

Accepted Answer

Positive-sounding words are used to mean the opposite

Answer

Sarcasm always uses very long words

Answer

Sarcasm only appears in other languages

Answer

Sarcasm has no words at all

Question 11

How can emojis like a smiley help a sentiment analysis program?

Accepted Answer

They give clear emotional clues about the writer's mood

Answer

They make the text load faster

Answer

They translate the text automatically

Answer

They hide the real sentiment

Question 12

What does 'polarity' refer to in sentiment analysis?

Accepted Answer

Whether text leans positive or negative, and how strongly

Answer

The number of letters in a word

Answer

The language a text is written in

Answer

Whether the text is a question

Question 13

What is a real-world use of sentiment analysis for a company that sells products online?

Accepted Answer

Automatically understanding if customers like or dislike products

Answer

Designing the company logo

Answer

Shipping the products faster

Answer

Setting the price of electricity

Question 14

What is the difference between 'binary' and 'multi-class' sentiment classification?

Accepted Answer

Binary uses two groups; multi-class uses three or more

Answer

Binary is for numbers; multi-class is for letters

Answer

Binary is faster computers; multi-class is slower ones

Answer

There is no difference

Question 15

In the 'bag of words' idea, what does the model mostly pay attention to?

Accepted Answer

Which words appear and how often, ignoring their order

Answer

Only the first letter of each word

Answer

The color and size of the text

Answer

The exact order of every word

Question 16

Why are very common words like 'the', 'is', and 'a' often removed before analysis?

Accepted Answer

They carry little feeling and crowd out meaningful words

Answer

They are spelled incorrectly

Answer

They are too long to process

Answer

They are in a foreign language

Question 17

Why might a sentiment model trained only on English movie reviews perform poorly on Hindi tweets?

Accepted Answer

It only learned patterns from English reviews, not Hindi tweets

Answer

Tweets cannot have any sentiment

Answer

Hindi has no positive words

Answer

Models always work on every language equally

Question 18

What is one ethical concern when using sentiment analysis on people's social media posts?

Accepted Answer

It can invade privacy if people did not agree to be analyzed

Answer

It uses too much paper

Answer

It makes the internet slower for everyone

Answer

It only works on weekends

Question 19

How can bias in training data cause an unfair sentiment model?

Accepted Answer

Unrepresentative data can make the model judge some groups unfairly

Answer

Bias makes the model run faster

Answer

Bias only affects the colors on screen

Answer

Bias has no effect on results

Question 20

What does a 'confidence score' tell you about a sentiment prediction?

Accepted Answer

How sure the model is about its prediction

Answer

How fast the model typed

Answer

How many words were in the text

Answer

The age of the model

Question 21

How is the 'accuracy' of a sentiment model usually measured?

Accepted Answer

By checking how many test examples it labels correctly

Answer

By counting how long it took to train

Answer

By measuring the file size of the model

Answer

By asking the model to grade itself

Question 22

Why should test data be kept separate from training data when building a model?

Accepted Answer

So the test fairly shows performance on new, unseen text

Answer

So the model can copy the answers

Answer

Because test data is always larger

Answer

To make training take longer

Question 23

What is 'subjectivity' in a piece of text, and how does it relate to sentiment?

Accepted Answer

It means the text expresses a personal opinion or feeling

Answer

It means the text is very long

Answer

It means the text has no verbs

Answer

It means the text is written by a computer

Question 24

In aspect-based sentiment analysis, what extra detail does the model find?

Accepted Answer

The sentiment about specific parts or features, not just overall

Answer

The exact time the review was written

Answer

How many people read the review

Answer

The price of the product

Question 25

Why does converting text to lowercase often help during preprocessing?

Accepted Answer

So the same word is counted together regardless of capital letters

Answer

Because lowercase letters are smaller files

Answer

Because computers cannot read capital letters

Answer

To make the text look prettier

Question 26

Which simple Python output would you expect from a basic sentiment tool given the text 'I really love this game'?

Accepted Answer

A positive label or a positive polarity score

Answer

A negative label

Answer

A spelling-error warning

Answer

The word count only

Question 27

In the tiny rule-based code below, what will it print for the review 'This food is terrible'?

positive = ['love','great','good']
negative = ['bad','terrible','hate']
review = 'This food is terrible'
if any(w in review for w in negative):
    print('negative')
elif any(w in review for w in positive):
    print('positive')
else:
    print('neutral')

Accepted Answer

negative

Answer

positive

Answer

neutral

Answer

an error message

Question 28

Why can misspelled words like 'amazng' lower a sentiment model's accuracy?

Accepted Answer

The model may not recognize the misspelled positive word

Answer

Misspelled words are always negative

Answer

Misspellings make the computer faster

Answer

Spelling never affects models

Question 29

Who usually creates the labels for a sentiment training dataset?

Accepted Answer

People (human annotators) read and tag the examples

Answer

The computer invents its own labels

Answer

The labels appear by magic

Answer

The internet provides them automatically

Question 30

How is sentiment analysis different from spam detection, even though both classify text?

Accepted Answer

Sentiment finds feeling; spam detection finds unwanted junk

Answer

They are exactly the same task

Answer

Sentiment analysis only works on emails

Answer

Spam detection measures word length

Test yourself