Wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

Introduction to Natural Language Processing (NLP)

Total questions: 102

Worksheet time: 51mins

Name
Class
Date
1.

Which statement best defines Natural Language Processing (NLP)?

a)

A field enabling computers to analyze human language

b)

A database system for storing large numeric datasets

c)

A method for compressing multimedia files for storage

d)

A protocol for networking devices across the internet

2.

Why is NLP important in computing?

a)

It encrypts all data using language models

b)

It increases screen resolution for sharper text

c)

It replaces binary with analog signals in hardware

d)

It converts human language into digital representations

3.

Which is a listed feature of natural languages?

a)

They are governed by syntax, lexicon, and semantics

b)

They always have fixed meanings that never change

c)

They are processed only through numerical vectors

d)

They avoid redundancy to prevent multiple meanings

4.

Which example illustrates autogenerated captions as an NLP application?

a)

Audio enhancement that removes background noise only

b)

Static subtitles manually typed by editors before release

c)

Live speech converted to on‑screen text during videos

d)

A slideshow presenting images with decorative font overlays

5.

Voice assistants use NLP primarily to do what?

a)

Understand spoken commands and execute tasks

b)

Encrypt conversations end‑to‑end for security

c)

Render 3D graphics for virtual environments

d)

Balance battery usage across mobile sensors

6.

Language translation in NLP is best described as

a)

Changing audio volume based on speech intensity

b)

Transforming pixels into words for screen readers

c)

Rewriting code comments into machine instructions

d)

Converting text or speech from one language to another

7.

What does sentiment analysis aim to detect in text?

a)

Whether content is positive, negative, or neutral

b)

How many users have read the content online

c)

Which author wrote the text and when

d)

The most frequent nouns across all documents

8.

A news website wants to sort articles into Food, Sports, and Politics. Which NLP tool fits this need?

a)

Text classification to assign predefined categories

b)

Named entity linking to external knowledge bases

c)

Topic modeling to discover hidden latent themes

d)

Dependency parsing to map grammatical relations

9.

Keyword extraction is mainly used to

a)

Measure reading speed of different users

b)

Generate entire paragraphs from short prompts

c)

Translate idioms between unrelated languages

d)

Automatically find important words and expressions

10.

Which statement about natural language redundancy is accurate?

a)

Redundancy should be removed from all languages

b)

It occurs only in informal speech, never in writing

c)

It prevents listeners from understanding context

d)

Multiple expressions can convey the same meaning

11.

Which scenario best demonstrates applying text classification in a practical workflow?

a)

Encrypting emails using public key certificates

b)

Improving microphone quality with noise suppression

c)

Compressing documents to reduce storage footprint

d)

Routing customer emails into support, billing, or sales

12.

A company tracks customer reviews to find emotions about its brand. Which NLP technique directly addresses this task?

a)

Automatic speech recognition to transcribe audio

b)

Machine translation to expand language coverage

c)

Sentiment analysis to judge polarity of opinions

d)

Keyword extraction to list frequent technical terms

13.

Which stage of NLP divides input into structural units like tokens within sentences and paragraphs?

a)

Lexical Analysis step of processing

b)

Syntactic Analysis of grammar

c)

Semantic Analysis for meaning

d)

Pragmatic Analysis for context

14.

During which NLP stage are grammatically incorrect constructions eliminated by checking relationships among words?

a)

Discourse Integration stage

b)

Semantic Analysis of sense

c)

Syntactic Analysis or parsing

d)

Lexical Analysis of tokens

15.

At what point does NLP ensure that phrases such as "hot ice cream" are rejected as meaningless?

a)

Semantic Analysis stage

b)

Pragmatic Analysis end

c)

Lexical Analysis at start

d)

Syntactic Analysis phase

16.

Which NLP stage checks intended meaning and context, often discarding literal meaning to match real-world use?

a)

Lexical tokenization step

b)

Semantic coherence check

c)

Discourse relation check

d)

Pragmatic interpretation stage

17.

What is the purpose of Discourse Integration in NLP?

a)

Builds token dictionary lists

b)

Forms story by linking sentences

c)

Maps words to lemma roots

d)

Parses clauses into trees

18.

Which statement best contrasts scriptbots and smart-bots?

a)

Scriptbots follow fixed scripts; smart-bots learn from data

b)

Scriptbots are flexible; smart-bots are limited

c)

Scriptbots require coding; smart-bots avoid programming

d)

Scriptbots need large databases; smart-bots use scripts

19.

Which trait is typical of a scriptbot?

a)

Wide functionality with learning

b)

Works around a programmed script

c)

Requires coding to adopt AI

d)

Uses bigger databases directly

20.

Which trait is typical of a smart-bot?

a)

Limited functionality overall

b)

No language processing skills

c)

Mostly free and easy to add

d)

Flexible and powerful design

21.

Why is text normalization used before deeper NLP tasks?

a)

To generate new training labels

b)

To convert raw text to canonical form

c)

To encrypt language for privacy

d)

To translate into human speech

22.

Which example illustrates normalization to a canonical form?

a)

Expanding text into paragraphs

b)

Replacing words with emojis

c)

Turning “goooood” into “good”

d)

Keeping all punctuation exactly

23.

Which sequence lists common normalization steps in plausible order?

a)

Parsing → discourse → pragmatic check

b)

Lemmatization → translation → parsing

c)

Segmentation → tokenization → stopword removal

d)

Stemming → tokenization → segmentation

24.

What does sentence segmentation accomplish in preprocessing?

a)

Splits corpus into sentences

b)

Maps words to vocabulary

c)

Removes stop words only

d)

Checks grammar relations

25.

In comparing human and computer language processing, which statement is accurate?

a)

Machines handle ambiguity easily

b)

Humans store information discretely

c)

Computers require numeric inputs

d)

Humans skip errors during typing

26.

Which chatbot capability most relies on Discourse Integration?

a)

Understanding a word list alone

b)

Maintaining coherence across turns

c)

Removing special characters only

d)

Converting text to lowercase

27.

Which step converts a sentence into separate tokens such as words and numbers?

a)

Tokenization process

b)

Lemmatization step

c)

TFIDF weighting

d)

Stemming process

28.

Which words are typically removed during preprocessing because they add little meaning to a sentence?

a)

All adjectives and adverbs

b)

Nouns like cat, tree

c)

Stop words like and, the

d)

Rare words like quark

29.

Why is text often converted to a common case (usually lower case) during preprocessing?

a)

To remove punctuation marks

b)

To avoid case‑sensitive mismatches

c)

To improve grammar rules

d)

To increase sentence length

30.

What does stemming primarily do to words?

a)

Finds word synonyms

b)

Strips affixes to root

c)

Counts term frequencies

d)

Predicts topic labels

31.

How does lemmatization differ from stemming?

a)

Removes punctuation only

b)

Counts words per document

c)

Produces meaningful lemmas

d)

Creates longer tokens

32.

In the example table, the lemma for studies is which word?

a)

studied

b)

studying

c)

student

d)

study

33.

What does a Bag of Words representation primarily capture from documents?

a)

Semantic role labels

b)

Word order patterns

c)

Syntactic parse trees

d)

Word occurrence counts

34.

Which is NOT a step in a typical Bag of Words pipeline?

a)

Create dictionary vocabulary

b)

Text normalization

c)

Parse dependency trees

d)

Build document vectors

35.

Term Frequency (TF) in a document is best described as which measure?

a)

Probability of a topic

b)

Inverse of document length

c)

Number of unique words

d)

Count of word occurrences

36.

Inverse Document Frequency (IDF) becomes small for which kind of word across a corpus?

a)

Very common stop words

b)

Words with many syllables

c)

Extremely rare terms

d)

Proper nouns only

37.

Which formula correctly defines IDF for a word W?

a)

Term count divided by document length

b)

TF multiplied by log of term count

c)

Total documents divided by documents with W

d)

Documents with W divided by total documents

38.

Which statement about TFIDF weighting is correct?

a)

TFIDF ignores term frequency entirely

b)

High TF and high IDF yield high score

c)

High TF and low IDF yield highest score

d)

Low TF and high IDF always zero score

39.

Which practical task commonly uses TFIDF in NLP applications?

a)

Speech synthesis voices

b)

Image edge detection

c)

Information retrieval ranking

d)

Graph shortest path search

40.

During stop word filtering, which of the following is likely to be removed from texts?

a)

data and model

b)

the and for

c)

vector and matrix

d)

random and sparse

41.

Given two documents, a word appears 5 times in one and 0 times in the other; it appears in only one of 100 documents in the corpus. Which option best describes its TFIDF behavior?

a)

Highest TFIDF in second document

b)

High TFIDF in first document

c)

Low TFIDF in first document

d)

Equal TFIDF in both documents

42.

Which statement best defines computer vision in the context of AI?

a)

A field enabling machines to derive meaning from visual data

b)

A tool for replacing all manual image editing tasks

c)

A method for compressing photos for faster storage

d)

A technique for increasing display screen resolution

43.

How does computer vision primarily differ from classical image processing?

a)

CV uses no machine learning, IP depends on deep models

b)

CV is a subset of IP with narrower applications

c)

CV only edits pixels, IP understands scene semantics

d)

CV interprets content for decisions, IP enhances image quality

44.

In the Emoji Scavenger Hunt example, what capability is demonstrated?

a)

Rendering 3D graphics for virtual environments

b)

Matching real objects to categories using camera input

c)

Encrypting photos before sharing with social apps

d)

Compressing video streams for low bandwidth

45.

Which is a correct pairing in the CV vs Image Processing table?

a)

CV example: object detection; IP example: rescaling brightness

b)

CV example: tone mapping; IP example: handwriting recognition

c)

CV main goal: enhance pixels; IP main goal: make predictions

d)

CV scope: subset of IP; IP scope: superset of CV

46.

Which application most directly uses facial features to verify identity?

a)

Facial recognition in attendance systems

b)

AR filters for playful face overlays

c)

Product search by uploaded photo

d)

Lane detection for autonomous cars

47.

What is the role of CV in smart homes and cities for security?

a)

Encrypt wireless traffic between sensors

b)

Generate synthetic voices for intercom systems

c)

Provide internet connectivity to household devices

d)

Recognize authorized individuals and track movements

48.

How do social media face filters operate in real time?

a)

Translate spoken words into text captions live

b)

Randomly place stickers without tracking features

c)

Detect and map facial landmarks, then overlay graphics

d)

Blur the background to reduce network usage

49.

Which outcome best illustrates Google’s Search by Image feature?

a)

Composes automatic captions from audio content

b)

Converts text to speech for accessibility features

c)

Encrypts uploaded images using secure hashing

d)

Returns results by matching visual patterns and shapes

50.

In retail, what does customer behavior tracking with CV aim to achieve?

a)

Optimize store layouts by analyzing movement paths

b)

Replace all human staff with autonomous robots

c)

Predict national economic trends from sales

d)

Prevent theft by encrypting video recordings

51.

Which statement about inventory management with CV is accurate?

a)

Images cannot indicate stock levels reliably

b)

Barcodes eliminate the need for any cameras onsite

c)

Cameras monitor shelves and trigger restock analysis

d)

Manual counting is faster than automated detection

52.

What CV tasks are essential for self-driving cars?

a)

Rendering movies at a higher frame rate

b)

Adjusting tire pressure automatically

c)

Streaming music based on driver preferences

d)

Detecting vehicles, pedestrians, signs, and lanes

53.

Which example reflects CV use in autonomous driving systems?

a)

Autopilot that detects nearby vehicles and maintains lanes

b)

Manual cruise control that fixes speed without sensors

c)

Navigation app that only computes shortest road distance

d)

Dashboard theme that changes colors with time

54.

How is CV applied in medical imaging to support diagnosis?

a)

Analyzing scans to detect abnormalities using models

b)

Watermarking images to protect hospital branding

c)

Replacing all radiologists with non-visual tests

d)

Compressing MRI files for quick email transfer

55.

What combination enables the Google Translate camera feature?

a)

RFID tags on signs provide multilingual codes

b)

LiDAR maps depth and GPS chooses language

c)

GANs generate fonts and NLP guesses grammar

d)

OCR detects text and AR overlays translated words

56.

Which CV task describes assigning a single label to an image?

a)

Classification for single-object recognition

b)

Reconstruction for building 3D surface models

c)

Segmentation for pixel-level contour labeling

d)

Tracking for temporal association of objects

57.

What does classification + localization add beyond classification?

a)

Identifies the object and its position in the image

b)

Generates synthetic training images automatically

c)

Predicts future object motion from past frames

d)

Converts 2D images into volumetric 3D models

58.

Which statement best distinguishes CV scope relative to image processing?

a)

CV only handles videos while image processing handles images

b)

CV is unrelated and excludes pixel-level operations

c)

CV is narrower than image processing in purpose

d)

CV is a superset that includes image processing tasks

59.

Which scenario shows CV improving the shopper experience?

a)

Cashier-less stores detecting items and charging accounts

b)

Printing paper coupons at manual checkout counters

c)

Only scanning barcodes with handheld devices

d)

Posting price labels without any sensor feedback

60.

Which statement best defines the pixel value in an 8-bit grayscale image?

a)

It is a float from 0 to 1 for hue

b)

It is an integer from 0 to 255 for intensity

c)

It is three integers storing red and blue

d)

It is a binary value indicating edges only

61.

In a digital image, what does resolution primarily indicate?

a)

The brightness range of the image

b)

The number of kernels used in CNNs

c)

The number of pixels composing an image

d)

The color depth per channel only

62.

Which correctly compares RGB and grayscale images?

a)

RGB stores one value per pixel, grayscale stores three

b)

Both store three channels but different ranges

c)

Both store one channel with different bit depths

d)

RGB stores three channel values, grayscale stores one

63.

In an 8-bit byte image, what do values 0 and 255 typically represent?

a)

0 is green, 255 is red

b)

0 is black, 255 is white

c)

0 is red, 255 is blue

d)

0 is white, 255 is black

64.

Which option lists common low-level visual features extracted from images?

a)

Layers, epochs, losses

b)

Edges, corners, blobs

c)

Weights, biases, neurons

d)

Windows, menus, icons

65.

What is a kernel in the context of convolution on images?

a)

A matrix that slides over pixels

b)

A random seed for training

c)

A color map for RGB display

d)

A loss function for classification

66.

Which best describes the convolution operation used in CNNs?

a)

Element-wise addition with random noise

b)

Element-wise multiplication with a sliding kernel

c)

Sorting pixel values across channels

d)

Averaging entire images into a scalar

67.

What is the primary role of the ReLU layer in a CNN?

a)

It normalizes features to mean zero only

b)

It converts RGB to grayscale before pooling

c)

It removes negative activations introducing non-linearity

d)

It reduces image resolution overall

68.

Which statement about pooling in CNNs is most accurate?

a)

Pooling increases feature map size

b)

Average pooling is identical to ReLU

c)

Pooling replaces kernels with larger strides

d)

Max pooling selects most prominent activations

69.

What does a fully connected layer do after convolution and pooling?

a)

Extracts edges using Sobel filters

b)

Flattens features and assigns labels

c)

Converts images into RGB channels

d)

Performs instance segmentation directly

70.

Which task requires assigning a label to each pixel of detected instances?

a)

Image classification only

b)

Object detection only

c)

Instance segmentation

d)

Feature extraction

71.

A 1280×1024 image is described as having high resolution because it has:

a)

Multiple color channels and depth

b)

Wide dynamic range of intensities

c)

Large kernels for convolution

d)

Many pixels arranged in rows and columns

72.

Which misconception about convolution is most plausible? Choose the correct statement.

a)

Convolution always blurs regardless of kernel

b)

Convolution sorts pixels by their coordinates

c)

Convolution highlights patterns via learned kernels

d)

Convolution ignores spatial neighborhoods entirely

73.

If an edge-detection filter is applied to a building photo, the output feature map will mainly show:

a)

Only the brightest windows

b)

Random noise across all pixels

c)

Boundaries outlining structures

d)

Uniform regions with constant color

74.

For multiple objects in an image, which pathway combines classification and localization?

a)

Histogram equalization process

b)

Clustering without labels

c)

Object detection for multiple items

d)

Classification only for single item

75.

Which statement best describes the hierarchical relationship among AI, ML, and DL?

a)

DL contains AI and ML within it

b)

AI contains ML and DL within it

c)

AI and ML are separate from DL entirely

d)

ML contains AI and DL within it

76.

Which option correctly contrasts rule-based and learning-based models?

a)

Rule-based use fixed rules; learning-based adapt from data

b)

Both require no data and only expert rules

c)

Rule-based adapt from data; learning-based use fixed rules

d)

Both adapt from data after manual labels

77.

In the Venn diagram of AI, ML, and DL, which circle is the smallest?

a)

Deep Learning circle

b)

Artificial Intelligence circle

c)

Machine Learning circle

d)

All circles are equal size

78.

Which example best fits a rule-based system?

a)

Chatbot with predefined question–answer pairs

b)

Spam filter that updates with new patterns

c)

Neural network learning from large datasets

d)

Image classifier trained on labeled data

79.

What is the primary input difference in the learning-based block shown?

a)

Unlabeled dataset is provided to model

b)

Only rules are provided to model

c)

Only answers are provided to model

d)

No data or rules are provided

80.

Supervised learning primarily relies on which type of data?

a)

Synthetic data generated randomly

b)

Noisy data without any labels

c)

Unlabeled data without any tags

d)

Labeled data provided for training

81.

Which task is classification rather than regression?

a)

Estimating tomorrow’s temperature

b)

Determining if an email is spam

c)

Forecasting a stock’s closing price

d)

Predicting a house’s sale price

82.

Which statement about deep learning from the notes is accurate?

a)

DL is a subset of ML using neural networks

b)

DL is broader than AI and includes ML

c)

DL replaces ML and rule-based systems

d)

DL uses rules instead of learning data

83.

Which outcome is expected from a rule-based approach block?

a)

Labels learned without any rules

b)

Clusters discovered from unlabeled data

c)

Answers produced from data plus rules

d)

Continuous values without fixed rules

84.

Which scenario best illustrates supervised learning in practice?

a)

Classifying images as dog or cat

b)

Tuning rules by human experts

c)

Grouping customers without labels

d)

Exploring data with no given targets

85.

Which distinction between regression and classification is correct?

a)

Both avoid using labeled training datasets

b)

Both predict only continuous numeric outputs

c)

Regression predicts classes; classification predicts numbers

d)

Regression predicts continuous values; classification predicts classes

86.

Why are learning-based models considered adaptive compared to rule-based models?

a)

They modify themselves using patterns in data

b)

They rely entirely on fixed human-written rules

c)

They require no data and only heuristics

d)

They output the same answers despite changes

87.

Which statement best describes unsupervised learning?

a)

Learns from labeled examples with targets provided

b)

Learns by memorizing training data without generalizing

c)

Learns on unlabeled data to find patterns and groups

d)

Learns via rewards from interacting with environment

88.

What is the main goal of clustering in unsupervised learning?

a)

Reduce dimensionality using principal components

b)

Map states to actions maximizing long-term returns

c)

Group similar data points into meaningful clusters

d)

Predict future numeric values from sequences

89.

Which scenario illustrates reinforcement learning most accurately?

a)

A model grouping shoppers by buying habits

b)

A model predicting salaries from past increments

c)

An agent learning to park via trial-and-error rewards

d)

A model labeling emails using provided categories

90.

In reinforcement learning, what provides learning signal to the agent?

a)

Human annotations for each data sample

b)

Random initialization of network weights

c)

Manual feature engineering by experts

d)

Positive and negative feedback as rewards

91.

Which pair correctly matches learning type with typical data?

a)

Association learning — labeled class targets

b)

Supervised learning — unlabeled data points

c)

Unsupervised learning — unlabeled data points

d)

Reinforcement learning — labeled image datasets

92.

Which best explains the difference between ANN and CNN?

a)

ANN cannot learn features; CNN has fixed filters

b)

ANN requires labels; CNN works only without labels

c)

ANN performs clustering; CNN performs classification only

d)

ANN models human neurons; CNN handles image features

93.

What is the primary advantage of artificial neural networks highlighted here?

a)

They eliminate the need for training data entirely

b)

They automatically extract features from raw inputs

c)

They guarantee perfect accuracy on large datasets

d)

They require manual rules for each decision

94.

Which layer in a neural network receives final processed information?

a)

Output layer

b)

Hidden layer

c)

Input layer

d)

Convolution layer

95.

What happens in the hidden layers of a neural network?

a)

Data is acquired without processing it

b)

Core processing using weights and biases occurs

c)

Labels are attached to each training example

d)

Predictions are displayed to the end user

96.

Which is a sub-category of unsupervised learning described in the material?

a)

Regression model

b)

Classification model

c)

Clustering model

d)

Temporal-difference model

97.

Which machine learning type is best when pre-existing data is insufficient but interaction feedback exists?

a)

Unsupervised learning

b)

Semi-supervised learning

c)

Supervised learning

d)

Reinforcement learning

98.

A supermarket wants to group customers by purchase patterns without prior labels. Which approach fits best?

a)

Association under reinforcement learning

b)

Clustering under unsupervised learning

c)

Q-learning under supervised learning

d)

Regression under supervised learning

99.

Which statement about CNNs aligns with the content?

a)

CNNs are unrelated to deep learning methods

b)

CNNs cannot differentiate objects within images

c)

CNNs use only fixed, untrained filters for edges

d)

CNNs assign learnable weights to image features

100.

Identify the correct sequence of layers in a simple feedforward neural network.

a)

Input → Hidden → Output

b)

Hidden → Input → Output

c)

Input → Output → Hidden

d)

Output → Hidden → Input

101.

Which description matches association models in unsupervised learning?

a)

Assign class labels using training examples

b)

Predict continuous outcomes from numeric inputs

c)

Discover co-occurrences among items bought together

d)

Map states to actions using cumulative rewards

102.

Which summary statement correctly matches ML model to purpose?

a)

Supervised: discover new patterns; Reinforcement: train with labels

b)

Unsupervised: discover patterns; Supervised: determine relationships via training

c)

Reinforcement: memorize data; Unsupervised: maximize reward signals

d)

Supervised: act in environments; Unsupervised: require labeled targets