Chapter: Natural Language Processing (NLP)

Q1. Fill in the blanks:

1. Data Science works around ______ and tabular data.

(numbers)
2. Computer Vision is focused on visual data like ______ and videos.
(images)
3. NLP stands for ______ Language Processing.
(Natural)
4. The sub-field of AI that focuses on understanding human languages is
______.
(NLP)
5. Natural Language Processing enables computers to process and analyze
large amounts of ______ data.
(natural language)
6. Automatic summarization helps in providing an overview of a ______ item
or blog post.
(news)
7. ______ analysis identifies sentiments and opinions from online content.
(Sentiment)
8. ______ categorization is used for spam filtering in emails.
(Text)
9. Google Assistant, Siri, and Alexa are examples of ______ powered by NLP.
(virtual assistants)
10. Speech recognition allows virtual assistants to detect and make sense of
______.
(speech)
11. The accuracy of a trained model is based on the ______ of the answers it
provides.
(relevance)
12. During evaluation, the chatbot’s suggested answers are compared to the ______
answers.
(actual)
13. The ______ line in the diagram represents the model’s output.
(blue)
14. The ______ line in the diagram represents the actual output.
(green)
15. ______ samples are used to evaluate the model’s efficiency.
(Data)
16. All the assistants like Google Assistant, Alexa, Cortana, Siri, etc. are considered as
__________.
(smart bots)
Q2. True/False Statements:
1. Data Science deals with visual data like images and videos.
False
2. NLP is a sub-field of AI that focuses on understanding natural languages.
True
3. Artificial Intelligence is unrelated to Linguistics.
False
4. NLP focuses on programming computers to analyze natural language data.
True
5. Mystery Animal game is an example of NLP.
True
6. Virtual assistants like Siri and Alexa use NLP for text classification.
False
7. Automatic summarization is relevant for avoiding redundancy in information.
True
8. Sentiment analysis helps companies understand customer opinions about their
products.
True
9. Text classification organizes documents into predefined categories.
True
10. Virtual assistants cannot access user data for tasks like sending messages.
False
11. The evaluation process does not require actual answers for comparison.
False
12. The blue line in the diagram refers to actual outputs.

False
13. Data samples are essential for understanding the efficiency of a model.
True
14. Accuracy is measured by comparing the chatbot’s answers with predefined
outputs.
True
15. Relevance of answers is not important for the evaluation of a chatbot.
False
16. Underfitting model’s accuracy is lower.
True
17. Overfitting model’s accuracy is high.
False
 18. Script bots are deployed in the customer care section of various companies.
True

(1 Mark Questions)
Q3. Expand the term?
a) NLP
b) CBT

Ans. a) Natural Language Processing

b) Cognitive Behavioural Therapy

Q4. What is the main purpose of automatic summarization in NLP?

Ans. To extract specific, important information from large datasets and provide an
overview while avoiding redundancy and maximizing content diversity.

Q5. Name one popular application of virtual assistants powered by NLP.

Ans. Google Assistant, Siri, Alexa, or Cortana.

Q6. What does text classification help achieve?

Ans. It assigns predefined categories to documents, organizes information, and simplifies

tasks, such as spam filtering in emails.

Q7. How is sentiment analysis useful for companies?

Ans. It helps companies understand customer opinions, assess their reputation, and gain
insights into purchasing decisions by analysing online sentiments.

Q8. Which technology allows virtual assistants to understand and respond to spoken
commands?

Ans. Speech recognition.

Q9. Define the term “Perfect fit”

Ans. A perfect fit occurs when a model matches well with the true function.

Q10. What are the functions of Script bots?

Ans. Their job is to answer some basic queries that they are coded for and connect them
to human executives once they are unable to handle the conversation.

Q11. What do you understand by corpus?

Ans. Whole textual data from all the documents altogether is known as corpus.
Q12. What does sentence segmentation mean?

Ans. Dividing the whole corpus into sentences is known as sentence segmentation.

Q13. What are tokens?

Ans. Tokens is a term used for any word or number or special character occurring in a
sentence.

Q14. What are stop words. Give some examples.

Ans. Stopwords are the words which occur very frequently in the corpus but do not add
any value to it.

For example: a, an, and are, the etc.

(2 Marks Questions)

Q1. How does sentiment analysis help companies understand customer feedback and
brand reputation?

Ans. Identifies emotions and opinions in text, helping companies understand customer
feedback and brand reputation.

Q2. Describe an application of text classification in everyday use.

Ans. One example is spam filters in emails. Text classification helps separate unwanted
emails (spam) from important ones, making email management easier.

Q3. Discuss the role of NLP in understanding emotional meaning from social media data.

Ans. NLP helps analyse feelings and opinions in social media posts. This allows
companies to understand how people feel about their brand or products and make
better decisions.

Q4. Write any four applications of NLP.

Ans. 1. Automatic Summarization:

2. Sentiment Analysis
3. Text Classification
4. Virtual Assistants

Q5. Define the term:

a) Overfitting
b) Underfitting

Ans.
a) Happens when a model learns the training data too well, including its noise, and
performs poorly on new data.
 b) Underfitting occurs when a model’s output does not match the true function at all.
Its accuracy is lower.

Q6. Name any four chatbots.

Ans. Mitsuku Bot, CleverBot, Haptik, Ochatbot

Q7. Write any two differences between Script bot and Smart bot.

Ans.
Script-bot Smart-bot
Script bots work around a script which is Smart bots work on bigger databases and
programmed in them. other resources directly
No or little language processing skills Coding is required to take this up on board
Limited functionality Wide functionality

Q8. Explain how automatic summarization helps in managing information overload.

Ans. Automatic summarization picks out the important information from large data to
extract key information, such as creating summaries of news articles or blogs.

Q9. How Stemming is different from Lemmatization? Explain with one example of each.

Ans. Stemming is a text normalization technique used in NLP to reduce words to their
root forms. It often results in incomplete or incorrect root forms
Example:
Word: "studies"
Stemmed form of word: "studi"

Lemmatization is also a text normalization technique used in NLP to reduce words

to their root forms. It results in meaningful word after removal of affix.
Example:
Word: "studies"
Stemmed form of word: "study"

Q10. Look at the diagram below and fill in the blanks:

1. _______________________

2. ______________________

Ans. 1. Lemmatization

2. Stemming
 Chapter: Evaluation
Q1. Fill in the blanks:
1. If Prediction and Reality both matches to Yes, then this condition is termed
as _____________.
True Positive
2. If Prediction and Reality both matches to No, then this condition is termed
as _____________.
True Negative
3. When Prediction is Yes and Reality is No. This condition is termed as
_____________.
False Positive
4. When Prediction is No and Reality is Yes. This condition is termed as
_____________.
5. False Negative

(1 Mark Questions)
Q3. Why confusion matrix is important?

Ans. The confusion matrix allows us to understand the prediction results.

Q2. When F1 Score is useful?

Ans. It is useful when both precision and recall are important, and there’s a need to
balance them.

Q3. What is the ideal value for the F1 Score, and what is the range of F1 Score values?

Ans. The ideal value for the F1 Score is 1 (100%), and the F1 Score ranges from 0 to 1.

(2 Marks Questions)

Q1. With reference to evaluation stage of AI project cycle, explain the term Precision.
Also give the formula to calculate it.

Ans. Precision is defined as the percentage of true positive cases out of all the cases
where the model predicted the positive class (i.e., true positives + false positives).
The formula for precision is:

Q2. With reference to evaluation stage of AI project cycle, explain the term Recall. Also
give the formula to calculate it.

Ans. Recall is the fraction of actual positive cases that are correctly identified by the
model.
Formula:
Q3. With reference to evaluation stage of AI, give any 2 real life scenarios when Precision
is More Important than Recall.

Ans. Precision is crucial when false positives (incorrectly predicting something as

positive) have serious consequences.

Example scenarios:
1. Spam detection: You don’t want to classify important emails as spam.
2. Medical diagnosis: Misdiagnosing a healthy person (false positive) as having a
disease can lead to unnecessary stress and expensive treatments.

Q4. With reference to evaluation stage of AI, give any 2 real life scenarios when Recall is
More Important than Precision.

Ans. Recall is important when false negatives (failing to detect positives) are more
harmful.

Example scenarios:
1. Cancer detection: Missing a cancer diagnosis (false negative) could lead to
severe consequences for the patient.
2. Fraud detection: Missing fraudulent activities can result in significant financial
losses.

Q5. With reference to evaluation stage of AI project cycle, explain the term F1 Score.
Also give the formula to calculate it.

Ans. F1 score is the measure of balance (harmonic mean) between precision and recall.
Formula:

(4 Marks Questions)

Q1. Suppose a model predicts the positive cases with the following outcomes:
• True Positives (TP): 30
• False Positives (FP): 10
• False Negatives (FN): 20
What will be the F1 score for the given problem?

Ans. Step 1: Precision

Step 2: Recall

Step 3: F1 Score

The F1 Score is 0.67 in this case, indicating a balanced performance of the model.
 ARTIFICIAL INTELLIGENCE
BOARD PAPER 2022

Section-A
(1 Mark Questions)

1. What is wage employment?

Ans. Wage employment is a type of employment where a person works for someone else
in exchange for wages or salary.

2. How food is one of the major problems related to sustainable development?

Discuss briefly.

Ans. Food insecurity, overproduction, unsustainable agricultural practices, and food waste
contribute to issues in sustainable development. Solutions include sustainable
farming, reducing waste, and equitable food distribution.

3. "To get success, every business idea needs to be unique or special." Is the
above statement a myth or fact?

Ans. Myth.
Explanation:
Success depends on execution, meeting customer needs, and market conditions,
not just uniqueness.

4. What is the major purpose of the Sustainable Development Goals?

Ans. The major purpose of SDGs is to eliminate poverty, reduce inequality, promote
sustainability, and ensure global prosperity and peace by 2030.

(2 Marks Questions)

5. Write any four common functions of an entrepreneur.

Ans. 1. Decision-making
An entrepreneur makes decisions what to produce or sell
2. Managing the Business
The entrepreneur plans for the future of his / her business
3. Taking Risk
Entrepreneurs take risks to protect the business from losses.
4. Innovation
Entrepreneurs innovate by trying new ideas or methods to increase income.

6. Sustainable development can actually happen only when each one of us works
towards it. Mention any four ways which we can do at our end to reduce
inequality.
Ans. 1. Promoting education for all
 2. Ensuring gender equality
3. Encouraging fair wages
4. Supporting marginalized communities

Section – B
(1 Mark Questions)

7. What is NLP?

Ans. Natural Language Processing (NLP) is a branch of AI that focuses on enabling

computers to understand, interpret, and respond to human language.

8. Mention any two commonly used applications of NLP.

Ans. 1. Chatbots,
2. Virtual assistants
3. Sentiment analysis
4. Machine translation.
(any two)

9. Name any two currently popular virtual assistants.

Ans. Alexa, Google Assistant, Siri, or Cortana. (any two)

10. Name the process of dividing whole corpus into sentences.

Ans. Sentence segmentation.

11. With reference to evaluation process of understanding the reliability of any AI

model, define the term True Positive.

Ans. True Positive (TP) refers to cases where the model correctly predicts a positive
outcome when it is actually positive.

12. What is F1 score?

Ans. The F1 score is the harmonic mean of precision and recall, providing a single metric
for evaluating model performance.

(2 Marks Questions)

13. Differentiate between Script-bot and Smart-bot.

Ans. Script-bot: Operates on predefined scripts with limited responses.

Smart-bot: Uses AI to understand and respond dynamically.
14. What is the purpose of Evaluation stage of AI project cycle? Discuss briefly.

Ans. The Evaluation stage assesses the performance and reliability of the AI model by
using metrics such as accuracy, precision, recall, and F1 score. It ensures the model
meets the desired objectives.

15. What is Tokenization? Count how many tokens are present in the following
statement:
I find that the harder I work, the more luck I seem to have.

Ans. Tokenization is the process of splitting text into smaller units, such as words or
sentences, called tokens.
Token Count:
Tokens: [I, find, that, the, harder, I, work, the, more, luck, I, seem, to, have]
Count: 14

16. Kaira, a beginner in the field of NLP is trying to understand the process of Stemming.
Help her in filling up the following table by suggesting appropriate affixes and stem
of the words mentioned there:

S. No. Word Affixes Stem

i. Tries
ii. Learning

Ans.
S. No. Word Affixes Stem
i. Tries -es Try
ii. Learning -ing Learn

Explanation:
• Tries: The affix -es is removed to derive the root word Try.
• Learning: The affix -ing is removed to derive the root word Learn.

17. With reference to evaluation stage of AI project cycle, explain the term Accuracy.
Also give the formula to calculate it.

Ans. Accuracy:
Accuracy is a metric used in the evaluation stage of the AI project cycle that measure
the percentage of correct predictions out of all the observations.

Formula:

or

where:
 • TP = True Positives
• TN = True Negatives
• FP = False Positives
• FN = False Negatives

18. Explain the following picture which depicts one of the processes on NLP.
Also mention the purpose which will be achieved by this process.

Ans. The image depicts the process of converting text to a common case, typically
lowercase, in NLP. This ensures that words with different capitalizations (e.g., "HELLO",
"HeLLo") are treated as the same, reducing case sensitivity.

Purpose: It normalizes the text, simplifies analysis, and improves machine

understanding by ensuring consistent word comparison.

Section - C
(COMPETENCY-BASED QUESTIONS)

(3 Marks Questions)

19. Consider the text of following documents:

Document 1: Sahil likes to play cricket
Document 2: Sajal likes cricket too
Document 3: Sajal also likes to play basketball

Apply all the four steps of Bag of words model of NLP on the above given documents
and generate the output.

Ans. Step 1: Tokenization

Document 1: [Sahil, likes, to, play, cricket]
Document 2: [Sajal, likes, cricket, too]
Document 3: [Sajal, also, likes, to, play, basketball]

Step 2: Vocabulary (Dictionary Creation)

Vocabulary: [Sahil, likes, to, play, cricket, Sajal, too, also, basketball]
 Step 3: Document Vectors (Frequency Table)
Document 1:
Sahil likes to play cricket Sajal too also basketball
1 1 1 1 1 0 0 0 0

Document 2:
Sahil likes to play cricket Sajal too also basketball
0 1 0 0 1 1 1 0 0

Document 3:
Sahil likes to play cricket Sajal too also basketball
0 1 1 1 0 1 0 1 1

Step 4: Document Vectors Table:

Sahil likes to play cricket Sajal too also basketball
1 1 1 1 1 0 0 0 0
0 1 0 0 1 1 1 0 0
0 1 1 1 0 1 0 1 1

20. With reference to NLP, explain the following terms in detail with the help of suitable
example:
• Term frequency
• Inverse Document Frequency

Ans. Term Frequency:

Term frequency is the frequency of a word in one document..
Formula:

Example: In Document 1 (Sahil likes to play cricket), TF of "cricket" = 1/5 = 0.2

Inverse Document Frequency:

Measures how important a term is across documents.

Formula:

Example: For "cricket," appearing in 2 out of 3 documents

IDF=log 3/2

21. Traffic Jams have become a common part of our lives now-a-days. Living in an urban
area means you have to face traffic each and every time you get out on the road.
Mostly, school students opt for buses to go to school. Many times, the bus gets late
due to such jams and students are not able to reach their school on time. Thus, an AI
model is created to predict explicitly if there would be a traffic jam on their way to
school or not. The confusion matrix for the same is
 The Confusion Matrix Actual: 1 Actual: 0
Predicted: 1 50 50
Predicted: 0 0 0

Explain the process of calculating F1 score for the given problem.

Ans.
 ARTIFICIAL INTELLIGENCE
BOARD PAPER 2023

Section-A
Objective Type Questions
(1 Mark Questions)

Q1. Answer the following questions.

i) Which of the following is not a task of an entrepreneur?
(a) Sharing of wealth
(b) Preferably using foreign materials
(c) Fulfilling customer needs
(d) Helping society

Ans. (b) Preferably using foreign materials

ii) GUI stands for:

(a) Graphical User Interaction
(b) Graphical User Interface
(c) Graphical Upper Interface
(d) None of these

Ans. (b) Graphical User Interface

iii) Which of the following is a function of an entrepreneur?

(a) Following the traditional method of business
(b) Innovation
(c) Keeping all the profit to himself/herself
(d) Avoid taking decisions

Ans. (b) Innovation

iv) Right clicking on File or Folder opens

(a) Main Menu
(b) Shortcut Menu
(c) Back Menu
(d) Front Menu

Ans. (b) Shortcut Menu

v) Stress management is vital because it leads to following benefits:

(a) Improves mood
(b) Boosts immune system
(c) Promotes longevity
(d) All of the above

Ans. (d) All of the above

 vi) Which of the following is an inner urge to do something, achieve their goals
without any external pressure lure for award or appreciation?
(a) Self-awareness
(b) Self-motivation
(c) Self-regulation
(d) Self-control

Ans. (b) Self-motivation

Q2. Answer the following questions.

i) Two popular examples of pocket assistants are ________ and ________.

Ans. Siri and Google Assistant

ii) This is a fact that all human beings have all nine types of intelligences, but at
different levels.
Name any two such intelligences.

Ans. Two types of intelligences are:

1. Linguistic Intelligence – The ability to use language effectively for
communication and expression.
2. Logical-Mathematical Intelligence – The ability to think logically, solve
mathematical problems, and understand abstract concepts.

iii) Identify the incorrect statements from the following:

(i) AI models can be broadly categorized into four domains.
(ii) Data sciences is one of the domains of AI model.
(iii) Price comparison websites are examples of data science.
(iv) The information extracted through data science can be used to
make decision about it.

(a) Only (iv)

(b) (iii) and (iv)
(c) Only (i)
(d) (ii) and (iii)

Ans. (d) (ii) and (iii)

Analysis of Each Statement

(i) AI models can be broadly categorized into four domains:
o This is correct. AI models are often categorized into domains such as
supervised learning, unsupervised learning, reinforcement learning, and
deep learning.

(ii) Data science is one of the domains of AI models:

o This is incorrect. While AI and data science are closely related fields, data
science is not a domain of AI; it is a separate interdisciplinary field that uses
AI techniques among other tools.
 (iii) Price comparison websites are examples of data science:
o This is incorrect. Price comparison websites are primarily applications of AI
(e.g., machine learning) rather than data science itself.

(iv) The information extracted through data science can be used to make
decisions about it:
o This is correct. Data science involves extracting insights from data, which are
then used to make informed decisions.

(iv) During Data Acquisition, feeding previous data into the machine is called:
(a) Training Data
(b) Predicting Data
(c) Testing Data
(d) Evaluating Data

Ans. (a) Training Data

(v) Regression is one of the type of supervised learning model, where data is
classified according to labels and data need not to be continuous.
(True/False)

Ans. False
Reason:
In regression, the output is continuous rather than discrete, unlike
classification where output is discrete.

(vi) Which of the following is defined as the measure of balance between

precision and recall?
(a) Accuracy
(b) F1 Score
(c) Reliability
(d) Punctuality

Ans. (b) F1 Score

Q3. Answer the given questions.

(i) Email filters, spam filters, smart assistants are the examples of:
(a) Pocket Assistants
(b) CV
(c) NLP
(d) Evaluation

Ans. (c) NLP (Natural Language Processing)

Explanation:
Email filters, spam filters, and smart assistants rely on NLP to process and
understand human language, making them examples of NLP applications.
 (ii) Select the correct features of Smart Bot:
(a) Smart-bots are flexible and powerful
(b) Coding is required to take this up on board
(c) Smart bots work on bigger databases and other resources directly
(d) All of the above

Ans. (d) All of the above

Explanation:
Smart-bots exhibit the following features:
• They are flexible and powerful, capable of handling various tasks
dynamically.
• Coding is often required for their implementation, especially for
customization.
• They can directly access and work on bigger databases and other
resources, making them robust and efficient.

(iii) For _______ the whole corpus is divided into sentences. Each sentence is
taken as a different data so now the whole corpus gets reduced to
sentences.
(a) Text Regulation
(b) Sentence Segmentation
(c) Tokenisation
(d) Stemming

Ans. (b) Sentence Segmentation

(iv) ___________ helps to find the best model that represents our data and
how well the chosen model will work in future.

Ans. Evaluation
Explanation:
Evaluation is the process that helps to determine the best model that fits the
data and assesses its performance for future predictions or tasks. It ensures
the model's reliability and efficiency.

(v) While evaluating a model's performance, recall parameter considers

____________.
(i) False positive
(ii) True positive
(iii) False negative
(iv) True negative
Choose the correct option:

(a) only (i)

(b) (ii) and (iii)
(c) (iii) and (iv)
(d) (i) and (iv)
 Ans. (b) (ii) and (iii)
Explanation:
The recall parameter measures the ability of a model to identify all relevant
instances (true positives). It is calculated using:

Thus, recall considers True Positives (TP) and False Negatives (FN).

(vi) With reference to NLP, consider the following plot of occurrence of words
versus their value:

In the given graph, X represents:

(a) Rare valuable words
(b) Punctuation words
(c) Popular words
(d) Pronoun

Ans. (a) Rare / valuable words

Explanation:
• Stop Words: High occurrence, low value (e.g., the, is, and).
• Frequent Words: Moderate occurrence, moderate value.
• Rare/Valuable Words (X): Low occurrence, high value, often carrying the
key meaning of the text.

Q4. Answer the given questions.

(i) Which of the following is a feature of document classification?
(a) Helps in classifying the type and genre of a document.
(b) Helps in creating a document.
(c) Helps to display important information of a corpus.
(d) Helps in including the necessary words in the text body.

Ans. (a) Helps in classifying the type and genre of a document.

(ii) Two conditions when prediction matches with the reality are true positive and
____________.
Ans. true negative.
 (iii) Which of the following is the correct feature of Neural network?
(a) It can improve the efficiency of two models.
(b) It is useful with small dataset.
(c) They are modelled on human brains and nervous system.
(d) They need human intervention.

Ans. (c) They are modelled on human brains and nervous system.

(iv) With reference to AI domain, expand the term CV.

Ans. Computer Vision

(v) Under __________, one looks at various parameters which affect the problem
we wish to solve, as this would make many lives better.

Ans. ethics

(vi) In this learning model, the data set which is fed to the machine is labelled.
Name the model.

Ans. Supervised Learning Model

Q5. Answer the given questions.

(i) _____________ is a term used for any word or number or special character
occurring in a sentence. (Token / Punctuator)

Ans. Token

(ii) When the prediction matches the reality, the condition is termed as
____________.

Ans. True Positive / True Negative

(iii) Smart Assistants such as Alexa, Siri are the examples of:
(a) Natural Language Processing
(b) Data Science
(c) Machine Learning
(d) Computer Vision

Ans. (a) Natural Language Processing

(iv) 4Ws Problem Canvas is a part of:

(a) Problem Scoping
(b) Data Acquisition
(c) Modelling
(d) Evaluation

Ans. (a) Problem Scoping

 (v) It refers to the unsupervised learning algorithm which can cluster the unknown
data according to the patterns or trends identified out of it.
(a) Regression
(b) Classification
(c) Clustering
(d) Dimensionality reduction

Ans. (c) Clustering

(vi) Which of the following talks about how true the predictions are by any model?
(a) Accuracy
(b) Reliability
(c) Recall
(d) F1 score

Answer: (a) Accuracy

Explanation:
Accuracy measures how many predictions made by the model are correct overall,
which directly relates to how true the predictions are.

Section - B
Subjective Type Questions
(2 Marks Questions)

Answer the questions on Employability Skills.

Q6. How does meditation help in Managing Stress? Discuss briefly.

Ans. Meditation helps in proper blood circulation, reducing anxiety and relaxing the body
which collectively reduce stress levels.

Q7. Give any two key roles performed by an entrepreneur.

Ans.
1. Innovation: Entrepreneurs identify new ideas and create products or services to
meet market demands.
2. Risk Management: Entrepreneurs take calculated risks to establish and grow their
businesses.

Q8. Mention any two benefits of working independently.

Ans.
1. Flexibility: Independent work allows individuals to set their own schedules and
work at their own pace.
2. Decision-making Power: Individuals have the freedom to make decisions without
external interference.

Q9. Gurmeet has just bought a new computer for his office. Suggest him any two points
which he should keep in mind to prevent his computer from virus infection.

Ans. (Any two points)

 1. Install an antivirus program and update it regularly.
2. Avoid downloading files or software from untrusted sources and do not open
suspicious email attachments.
3. Use passwords to login to your computer
4. Use secure sites

Q10. Define the term agricultural entrepreneurship. How are farmers benefitted from it?

Ans. Agricultural entrepreneurship refers to the application of innovative business

practices in agriculture to improve productivity, profitability, and sustainability.

Benefits to farmers:
1. Increased Income: Adoption of modern techniques and value-added products
increases profitability.
2. Market Access: Entrepreneurs help farmers connect directly with markets,
reducing dependency on intermediaries.

Answer the given questions in 20-30 words each.

Q11. Explain any one example of AI bias.

Ans. AI bias occurs when systems produce unfair results.

For example, an AI hiring tool may favour male candidates over female due to biased
training data.

Q12. What is Dimensionality Reduction?

Ans. Dimensionality reduction is the process of reducing the number of features in a

dataset while retaining essential information to improve efficiency and reduce
computational complexity.

Q13. Define Chatbot. What are its types?

Ans. A chatbot is an AI-powered tool that simulates human conversation.

Types:
1. Script-bot (Rule-based Chatbots) (respond to predefined commands).
2. Smart-bot (AI Chatbots) (learn from user interactions).

Q14. Define Confusion Matrix.

Ans. A confusion matrix is a table that records the result of comparison between the
prediction and reality.

Q15. Face lock feature of a smartphone is an example of computer vision. Briefly discuss
this feature.

Ans. The face lock feature uses computer vision to identify facial patterns via camera,
comparing them with stored biometric data to authenticate and unlock the
smartphone.
Q16. With reference to data processing, expand the term TFIDF. Also give any two
applications of TFIDF.

Ans. TFIDF stands for Term Frequency-Inverse Document Frequency.

Applications: (any two only)
1. Text Summarization (identifying key terms in documents).
2. Search Engines (Information Retrieval).
3. Stop word filtering (removal of unwanted words out of text)

(4 Marks Questions)
Answer the given questions in 50-80 words each.

Q17. Ms. Sooji is a beginner in the field of Artificial Intelligence. She got confused among
the core terms like Artificial Intelligence (AI), Machine Learning (ML) and Deep
Learning (DL). Many a times, these terms are used interchangeably but are they the
same? Justify your answer. Help her in understanding these terms by drawing a well
labelled diagram to depict the interconnection of these three fields.

Ans. 1. Artificial Intelligence (AI):

AI refers to techniques that enable computers to mimic human intelligence by
performing tasks such as reasoning, decision-making, and problem-solving.

2. Machine Learning (ML):

ML is a subset of AI where machines learn from data to make predictions or
decisions without explicit programming.

3. Deep Learning (DL):

DL is a subset of ML that uses artificial neural networks with multiple layers to
process large amounts of data and solve complex problems like image
recognition, natural language processing, and autonomous driving.

AI

ML

DL

Q18. What is the significance of AI project cycle? Also explain in detail about how Data
Acquisition is different from data exploration.

Ans. The AI project cycle provides a structured approach for building AI systems, ensuring
efficient and effective solutions.
 Data Acquisition involves collecting raw data from various sources, while Data
Exploration focuses on understanding the data by cleaning, visualizing, and analysing
it for patterns or anomalies. Data exploration ensures the data is ready for modelling,
making it critical for AI project success.

Q19. Create a document vector table from the following documents by implementing all
the four steps of Bag of words model. Also depict the outcome of each step.
Document 1: Sameera and Sanya are classmates.
Document 2: Sameera likes dancing but Sanya loves to study mathematics.

Ans. Step 1: Tokenization

Document 1: [Sameera, and, Sanya, are, classmates]
Document 2: [Sameera, likes, dancing, but, Sanya, loves, to, study, mathematics]

Step 2: Vocabulary (Dictionary Creation)

Vocabulary: [Sameera, and, Sanya, are, classmates, likes, dancing, but, loves, to,
study, mathematics]

Step 3: Document Vector

Document 1:
Sameera and Sanya are classmates likes dancing but loves to study mathematics
1 1 1 1 1 0 0 0 0 0 0 0

Document 2:
Sameera and Sanya are classmates likes dancing but loves to study mathematics
1 0 1 0 0 1 1 1 1 1 1 1

Step 4: Document Vectors Table

Sameera and Sanya are classmates likes dancing but loves to study mathematics
1 1 1 1 1 0 0 0 0 0 0 0
1 0 1 0 0 1 1 1 1 1 1 1

OR
Sameera 1 1
and 1 0
Sanya 1 1
are 1 0
classmates 1 0
likes 0 1
dancing 0 1
but 0 1
loves 0 1
to 0 1
study 0 1
mathematics 0 1
Q20. Will it be valid to say that not all the devices which are termed as "smart" are AI-
enabled? Justify this statement. Explain any two examples from the daily life which are
commonly misunderstood as AI.

Ans. Not all "smart" devices are AI-enabled, as some rely on pre-programmed responses
or basic automation rather than intelligent learning. For instance:

1. A fully automatic Washing Machines: It works on its own, but it requires human
intervention to select the parameters of washing.
2. Wi-Fi enabled AC: It can be turned on and off remotely with the help of internet
but does not have decision-making abilities. This is an example of Internet of
Things (IoT)

Q21. Recently the country was shaken up by a series of earthquakes which has done a huge
damage to the people as well as the infrastructure. To address this issue, an AI model
has been created which can predict if there is a chance of earthquake or not. The
confusion matrix for the same is:

Reality
Confusion Matrix
Yes No
Yes 50 05
Predicted
No 25 20

(i) How many total cases are True Negative in the above scenario?
(ii) Calculate precision, recall and F1 score.

Ans. (i) True Negatives: 20 cases

(ii)
 ARTIFICIAL INTELLIGENCE
BOARD PAPER 2024

Section-A
Objective Type Questions
(1 Mark Questions)

1. Answer the given questions:

(i) SMART method can be used to set goals to make you successful in your career
and personal life. What does 'A' in SMART stand for?
(a) Abrupt
(c) Achievable
(b) Accountable
(d) Admirable

Ans. (c) Achievable

(ii) Which of the following is not a key element of self-management skills?

(a) Prioritising your work
(b) Not taking feedback
(c) Goal setting
(d) Staying updated about new practices

Ans. (b) Not taking feedback

(iii) Which of the following is a quality of successful entrepreneurs?

(a) Hard working
(b) Resistance to change
(c) Lazy
(d) Less-confident

Ans. (a) Hard working

(iv) The most important software in any computer is the ________. This is the
software that starts working as soon as we switch on a computer.
(a) Web Browsers
(b) Operating System
(c) Office Software
(d) Designing Software

Ans. (b) Operating System

(v) Which of the following types of communication takes place when one
individual address a large gathering?
(a) Written communication
(b) Public communication
(c) Small group communication
(d) Interpersonal communication
 Ans. (b) Public communication

(vi) Assertion (A) : Organic farming technique is an example of a green skill that is
essential for sustainable agriculture.
Reason (R): Organic farming technique prioritise environment friendly and
sustainable practices such as using natural fertilisers, avoiding synthetic
pesticides, and promoting soil health.
(a) Both (A) and (R) are true and (R) is the correct explanation for (A).
(b) Both (A) and (R) are true and (R) is not the correct explanation of (A).
(c) (A) is true, but (R) is false.
(d) (A) is false, but (R) is true.

Ans. (a) Both (A) and (R) are true, and (R) is the correct explanation for (A).

2. Answer the given questions:

(i) Which of the following contributes to the efficiency of an AI project?
(a) High Model Complexity
(b) Relevant and Authentic Training Data
(c) Minimal Preprocessing
(d) Limited Hardware Resources

Ans. (b) Relevant and Authentic Training Data

(ii) This real-life application of NLP is used to provide an overview of a news item
or blog post, while avoiding redundancy from multiple sources and
maximising the diversity of content obtained. Which is this application?
(a) Chatbot
(b) Virtual Assistant
(c) Sentiment Analysis
(d) Automatic Summarisation

Ans. (d) Automatic Summarisation

(iii) Which of the following represent a machine that is smart but not considered
Artificial Intelligence (AI) enabled?
(a) A robotic vacuum cleaner that can navigate and clean floors
autonomously.
(b) A chatbot that engages in natural language conversations and answers
questions.
(c) A smartphone with facial recognition for unlocking the device.
(d) A digital alarm clock that rings at a set time every morning.

Ans. (d) A digital alarm clock that rings at a set time every morning.

(iv) Which of the following words represent an example of a lemma resulting from
lemmatisation for "caring" in context to Natural Language Processing (NLP)?
(a) Care
(b) Cared
 (c) Cares
(d) Car

Ans. (a) Care

(v) Intrapersonal Intelligence is a concept that:

(a) Measures an individual's ability to understand others' emotions and
feelings.
(b) Assesses one's proficiency in mathematics and logical reasoning.
(c) Describes the level of self-awareness someone has, starting from realizing
weaknesses, strengths, to recognizing their own feelings.
(d) Evaluates an individual's spatial navigation and visualisation skills.

Ans. (c) Describes the level of self-awareness someone has, starting from realizing
weaknesses, strengths, to recognizing their own feelings.

(vi) For Data Science, usually the data is collected in the form of tables. These
tabular datasets can be stored in different formats. Which of the following
formats is not used for storing data in a tabular format?
(a) CSV
(b) Website
(c) SQL
(d) Spreadsheet

Ans. (b) Website

Explanation:

• CSV (Comma-Separated Values): A standard file format for storing tabular

data.
• SQL (Structured Query Language): Used to store and manage data in
relational database tables.
• Spreadsheet: Applications like Excel or Google Sheets store data in a
tabular format.
• Website: While websites can display tabular data, they are not inherently
a storage format for tabular data. Websites are typically used to present
information, not to store structured data in tabular form.

3. Answer the given questions:

(i) _________ is one of the parameters for evaluating a model's performance and
is defined as the fraction of positive cases that are correctly identified.
(a) Precision
(b) Accuracy
(c) Recall
(d) F1

Ans. (c) Recall

(ii) In the AI project cycle, which of the following represents the correct order of
steps?
(a) Data Exploration, Problem Scoping, Modelling, Evaluation, Data
Acquisition.
(c) Problem Scoping, Data Acquisition, Data Exploration, Modelling,
Evaluation.
(d) Modelling, Data Acquisition, Evaluation, Problem Scoping, Data
Exploration.
(e) Data Acquisition, Data Exploration, Problem Scoping, Modelling,
Evaluation.

Ans. (c) Problem Scoping, Data Acquisition, Data Exploration, Modelling, Evaluation

(iii) __________ is a concept to unify statistics, data analysis, machine learning and
their related methods in order to understand and analyse actual phenomena
with data.
(a) Computer Vision
(b) Natural Language Processing
(c) Data Science
(d) Computer Science

Ans. (c) Data Science

(iv) In computer vision which of the following tasks is used for multiple objects?
(a) Classification
(b) Classification + Localisation
(c) Instance Segmentation
(d) Localisation

Ans. (c) Instance Segmentation

(v) In spam email detection, which of the following will be considered as"False
Negative" ?
(a) When a legitimate email is accurately identified as not spam.
(b) When a spam email is mistakenly identified as legitimate.
(c) When an email is accurately recognised as spam.
(d) When an email is inaccurately labelled as important.

Ans. (b) When a spam email is mistakenly identified as legitimate.

(vi) Which of the following applications is not associated with Natural Language
Processing (NLP)?
(a) Sentiment Analysis
(b) Speech Recognition
(c) Spam Filtering in emails
(d) Stock Market Analysis

Ans. (d) Stock Market Analysis

4. Answer the given questions:
(i) Statement 1: Confusion matrix is an evaluation metric.
Statement 2: Confusion Matrix is a record which helps in evaluation.
(a) Both Statement 1 and Statement 2 are correct.
(b) Both Statement 1 and Statement 2 are incorrect.
(c) Statement 1 is correct and Statement 2 is incorrect.
(d) Statement 2 is correct and Statement 1 is incorrect.

Ans. (a) Both Statement 1 and Statement 2 are correct.

(ii) Which form of unsupervised learning does the following diagram indicate?

(a) Clustering
(b) Regression
(c) Reinforcement learning
(d) Classification

Ans. (a) Clustering

(iii) Bag of Words is a model which helps in extracting features out of the text which
can be helpful in machine learning algorithms.
(a) Data Science (DS)
(b) Virtual Reality (VR)
(c) Natural Language Processing (NLP)
(d) Computer Vision (CV)

Ans. (c) Natural Language Processing (NLP)

(iv) Which of the following represents an example of a recommendation system?

(a) An online clothing store that offers a wide variety of clothing options.
(b) A search engine that retrieves relevant web pages based on user queries.
(c) An e-commerce website that displays customer reviews and ratings for
products.
(d) A music streaming platform that suggests songs and playlists based on user
listening history.

Ans. (d) A music streaming platform that suggests songs and playlists based on user
listening history.

(v) Name any two search engines.

Ans. Google and Bing
 (vi) What is the primary need for evaluating an AI model's performance in the AI
Model Development process?
(a) To increase the complexity of the model.
(b) To visualize the data.
(c) To assess how well the chosen model will work in future.
(d) To reduce the amount of data used for training.

Ans. (c) To assess how well the chosen model will work in future.

5. Answer the given questions:

(i) Assertion (A): The term used to refer to the number of pixels in an image is
resolution.
Reason (R): Resolution in an image denotes the total number of pixels it
contains, usually represented as height x width.
(a) Both (A) and (R) are true and (R) is the correct explanation for (A).
(b) Both (A) and (R) are true and (R) is not the correct explanation for (A).
(c) (A) is true, but (R) is false.
(d) (A) is false, but (R) is true.

Ans. (a) Both (A) and (R) are true, and (R) is the correct explanation for (A).

(ii) When a machine possesses the ability to mimic human traits, i.e., make
decisions, predict the future, learn, and improve on its own, it is said to have:
(a) Computational Skills
(b) Learning Capability
(c) Artificial Intelligence
(d) Cognitive Processing

Ans. (c) Artificial Intelligence

(iii) Statement 1: To evaluate a models' performance, we need either precision or

recall.
Statement 2: When the value of both Precision and Recall is 1, the F1 score is
0.
(a) Both statement 1 and statement 2 are correct.
(b) Both statement 1 and statement 2 are incorrect.
(c) Statement 1 is correct, but statement 2 is incorrect.
(d) Statement 1 is incorrect, but statement 2 is correct.

Ans. (c) Statement 1 is correct, but Statement 2 is incorrect.

Explanation for Statement 2:
When precision and recall are both 1, the F1 score is also 1 (not 0) because it
is the harmonic mean of precision and recall.

(iv) The concept of ________ is used to apply face filters on various social media
platforms.
(a) NLP
 (b) Computer Vision
(c) Data Science
(d) Block chain Technology

Ans. (b) Computer Vision

(v) The 4 W's Problem Canvas helps in identifying the key elements related to the
given problem.
Which of the following is NOT one of the blocks of the Problem Canvas?
(a) When
(b) Where
(c) What
(d) Why

Ans. (a) When

(vi) Which domain of AI is used for interacting with virtual assistants such as Siri
and Alexa?
(a) Machine Learning (ML)
(b) Computer Vision (CV)
(c) Natural Language Processing (NLP)
(d) Technical Vision (TV)

Ans. (c) Natural Language Processing (NLP)

Section B
(Subjective Type Questions)

Answer the questions on Employability Skills. Answer each question in 20-30

words.
6. Give any two examples of how individual choices and behaviours can contribute in
achieving sustainable development.

Ans. 1. Reducing energy consumption by using energy-efficient appliances.

2. Minimizing waste by practicing recycling and reusing materials.

7. List any two common misconceptions about entrepreneurship.

Ans. 1. Every business idea needs to be unique or special.

2. Entrepreneurs are born, not made.
3. A person needs a lot of money to start a business.
4. A person having a big business is an entrepreneur.
(Any two)

8. What is the importance of time management in effectively dealing with stress? Provide
any one strategy for improving time management skills to reduce stress.
Ans. Time management helps prioritize tasks, avoiding last-minute pressure and reducing
stress.
Strategy: Use a to-do list to organize tasks and set deadlines.

9. Mention any two measures that individuals or organisations can take to protect their
data from theft and viruses.

Ans. 1. Use strong passwords and regularly update it.

2. Install antivirus software and update it regularly.

10. The method of communication that you choose could affect the relationship with your
peers, superiors, and customers. Write the four factors on the basis of which you can
choose the right method of communication.

Ans. 1. Urgency of the message.

2. Nature of the message (formal or informal).
3. Relationship with the recipient.
4. Accessibility of the recipient.

Answer any 4 out of given 6 questions in 20-30 words each.

11. Differentiate between Machine Learning (ML) and Deep Learning (DL).

Ans. 1. ML is a subset of AI where machines learn from data to make predictions or

decisions without explicit programming.
2. DL is a subset of ML that uses artificial neural networks with multiple layers to
process complex tasks.

12. What are the primary differences between Script-bots and Smart-bots?

Ans. 1. Script-bots works around pre-written commands with no adaptability.

2. Smart-bots use AI to learn, adapt, and respond dynamically.

13. What do you mean by Evaluation of an AI model? Also explain the concept of
overfitting with respect to AI model Evaluation.

Ans. Evaluation is the process to measure an AI model's accuracy and performance.

Overfitting occurs when a model performs well on training data but poorly on but
poorly on unseen testing data

14. For a healthcare organisation's objective of predicting disease outbreaks and

efficiently allocating resources through the analysis of medical records, would you
recommend using supervised learning or unsupervised learning as the preferred
machine learning approach? Explain your choice briefly.

Ans. Supervised learning is preferred as it uses labelled data to predict disease outbreaks
accurately, based on medical records.
15. What role does data play in AI based applications? Name any two sources of online
data collection for building any AI based application.

Ans. Data is used to train AI models to identify patterns and make decisions.
Sources: APIs (e.g., Twitter API), Open datasets (e.g., Kaggle), Open-sourced
Government Portals (any two)

16. Differentiate between grayscale and RGB images.

Ans. 1. Grayscale images use shades of Gray, with one intensity channel.
2. RGB images combine red, green, and blue channels to display colours.

Answer the given questions in 50-80 words each.

17. What are Neural networks? Briefly explain all the layers of a neural network.

Ans. Neural networks are computational models inspired by the human brain, used for
tasks like classification and regression.
• Input Layer: Receives raw data for processing.
• Hidden Layer(s): Perform calculations and extract features.
• Output Layer: Produces the result or prediction.

18. Give any four examples of applications of AI that we see around us.

Ans. 1. Virtual assistants like Siri and Alexa.

2. Recommendation systems on Netflix and Amazon.
3. Face recognition in smartphones.
4. Autonomous vehicles like self-driving cars.

19. Consider the following two documents:

Document 1: ML and DL are part of AI.
Document 2: DL is a subset of ML.
Implement all four steps of the Bag of Words (BoW) model to create a document
vector table. Depict the outcome of each step.

Ans. Step 1: Tokenization: Break documents into words.

Doc 1: ["ML", "and", "DL", "are", "part", "of", "AI"]
Doc 2: ["DL", "is", "a", "subset", "of", "ML"]

Step 2: Vocabulary (Dictionary Creation)

["ML", "and", "DL", "are", "part", "of", "AI", "is", "a", "subset"]

Step 3: Frequency Calculation (Document Vector):

Document 1:
ML and DL are part of AI is a subset
1 1 1 1 1 1 1 0 0 0
 Document 2:
ML and DL are part of AI is a subset
1 0 1 0 0 1 0 1 1 1

Step 4: Document Vector Table

ML and DL are part of AI is a subset
1 1 1 1 1 1 1 0 0 0
1 0 1 0 0 1 0 1 1 1

OR
Word Doc 1 Doc 2
ML 1 1
and 1 0
DL 1 1
are 1 0
part 1 0
of 1 1
AI 1 0
is 0 1
a 0 1
subset 0 1

20. Consider the following graphs (Figure 1 and Figure 2) that demonstrate the two types
of Supervised Learning Models of Artificial Intelligence. Identify and explain each
model giving suitable examples of each.

Ans. Figure 1:
Classification Model: Categorizes data into distinct classes, like spam detection or
image recognition.
Figure 2:
Regression Model: Predicts continuous values, such as house prices or stock market
trends.

21. A binary classification model has been developed to classify news articles as either
"Fake News" or "Real News". The model was tested on a dataset of 500 news articles,
and the resulting confusion matrix is as follows:

Reality
Confusion Matrix
Yes No
Yes 45 15
Predicted
No 20 420
 (A) How many total cases are True Negative in the above scenario?
(B) Calculate Precision, Recall and F1-Score.

Ans.