MODULE – 1

1. Define Artificial Intelligence. Explain its scope.

Definition of AI: Artificial Intelligence (AI) refers to the simulation of human intelligence in
machines that are programmed to think and learn like humans. It enables machines to
perform tasks such as:

• Learning from experience (machine learning)

• Understanding language (NLP)

• Making decisions

• Recognizing patterns (vision and speech)

• Solving problems logically

These machines can adapt to new inputs and perform human-like tasks, often faster and with
greater precision.

Scope of AI:
AI has widespread applications across many fields. Its scope includes:

• Problem Solving: AI can tackle complex problems, such as finding the best route on
a map (e.g., Google Maps), detecting fraud in banking, or diagnosing diseases in
healthcare.

• Natural Language Processing (NLP): AI powers applications that understand and

generate human language, such as chatbots, translators, or virtual assistants like Alexa
and Siri.

• Robotics: Robots use AI to perform tasks like cleaning (Roomba), delivery

(autonomous drones), or even performing surgeries.

• Computer Vision: AI allows machines to “see” and interpret visual data — used in
face detection, medical imaging, or quality inspection in factories.

• Expert Systems: These simulate a human expert's decision-making, commonly used

in healthcare (e.g., MYCIN) or legal analysis.

• Machine Learning: AI systems learn patterns from data and improve their
performance, used in spam filtering, recommendation engines (like Netflix), or stock
market predictions.

• Speech Recognition: Converts spoken language into text (used in voice-to-text apps
or voice commands).
• Industry Applications:
 o Healthcare – Diagnosis, drug discovery

o Finance – Risk analysis, fraud detection

o Education – Personalized learning

o Agriculture – Crop monitoring, smart irrigation

o Manufacturing – Predictive maintenance, automation

2. Explain the history and evolution of Artificial Intelligence.

Early Days (1940s–1950s):

• Alan Turing proposed the concept of machines that could simulate human thinking.

• In 1950, he developed the Turing Test, which checks if a machine can exhibit
intelligent behavior indistinguishable from a human.

• 1956: The Dartmouth Conference, organized by John McCarthy, officially marked

the beginning of AI as a research field.

Initial Progress (1956–1970):

• Simple AI programs like:

o Logic Theorist – Proved mathematical theorems.

o General Problem Solver (GPS) – Solved a wide range of problems using
heuristics.

• These programs operated on symbolic logic and predefined rules.

AI Winter (1970s–1980s):

• Hype and overpromises led to disappointment.

• AI faced criticism for not delivering practical results.

• Funding and research declined.

Expert Systems Era (1980s):

• Systems like MYCIN used rules to diagnose diseases.

• AI found real-world applications in medicine, industry, and logistics.

• Limited adaptability — systems couldn’t learn or handle uncertainty.

Machine Learning & Big Data Era (1990s–2000s):

• Shift from rules to learning from data.

• Introduction of algorithms that could improve with experience.
 • Rise of big data and internet helped train models more effectively.

Modern AI (2010–Present):

• Breakthroughs in deep learning, neural networks, and reinforcement learning.

• Applications: Self-driving cars, chatbots, virtual assistants, facial recognition.

• Tools: ChatGPT, Google Assistant, Siri, Alexa.

• AI now powers most digital platforms and smart devices.

3. Discuss the types of Artificial Intelligence.

Based on Capability:

1. Narrow AI (Weak AI):

o Performs a specific task only.

o Cannot generalize beyond its programmed purpose.
o Examples: Face recognition, spam filters, voice assistants.

2. General AI (Strong AI):

o Can understand, learn, and apply knowledge like a human across a wide range
of tasks.

o Still theoretical and under development.

3. Super AI:

o Hypothetical future AI that surpasses human intelligence.

o Would have self-awareness and the ability to outperform humans in all fields.

o Raises ethical and safety concerns.

Based on Functionality:

1. Reactive Machines:

o No memory, only react to current input.

o Example: IBM’s Deep Blue chess-playing computer.

2. Limited Memory:

o Remembers past actions temporarily to inform decisions.

o Example: Self-driving cars analyzing previous traffic data.

3. Theory of Mind:
 o Can understand human emotions, beliefs, and intentions.

o Still under research.

4. Self-Aware AI:

o Has its own consciousness, emotions, and self-awareness.

o Theoretical at this stage.

4. Differentiate between Narrow AI and General AI

Feature Narrow AI General AI

Capability Performs a specific task Performs any intellectual task

Flexibility One domain only Multiple domains

Development Already developed and used widely Still in research

Examples Chatbots, recommendation systems Human-like robots (hypothetical)

Learning Task-specific learning Universal learning and application

5. Explain problem formulation and problem-solving techniques in AI.

Problem Formulation: It's about defining the problem in a way that an AI system can
understand and process.

A problem consists of:

• Initial state – Where we start

• Actions – Possible moves

• Transition model – What happens after an action

• Goal state – Desired outcome

• Path cost – Cost associated with reaching the goal

Problem-Solving Techniques:

1. Search-Based Techniques:
o Explore different paths to reach the goal.

o Example: Maze solving, route finding.

2. Knowledge-Based Techniques:

o Use logic and facts to derive conclusions.

 o Example: Expert systems using "if-then" rules.

3. Heuristics:

o Approximate methods to guide the search efficiently.

o Based on intuition or experience.

6. What are search algorithms in AI? Explain uninformed and informed search
strategies.
Search Algorithms: Used by AI to explore all possible states and find a solution. They help
find a path from the initial state to the goal state.
Uninformed (Blind) Search:

• No extra information is used.

• Explore all possibilities blindly.

Examples:

• Breadth-First Search (BFS): Explores all nodes level by level.

• Depth-First Search (DFS): Explores one path deeply before trying others.

• Uniform Cost Search: Chooses the lowest cost path at each step.
Informed (Heuristic) Search:

• Uses domain-specific knowledge (heuristics) to make smarter choices.

Examples:

• Greedy Best-First Search: Chooses node that seems closest to goal.

• A Search:* Uses both path cost and estimated cost to goal:

o f(n) = g(n) + h(n)

o g(n) = cost so far, h(n) = estimated cost to goal.

7. What is heuristic search? Explain constraint satisfaction problems (CSP) in detail.

Heuristic Search:

• A problem-solving technique that uses rules or estimates (heuristics) to make

decisions faster.

• Heuristics help prioritize which paths to explore first.

Examples of Heuristic Search:
 • A* – Most popular pathfinding algorithm.

• Hill Climbing – Keeps moving toward higher value (like climbing a hill).

• Best-First Search – Always picks the most promising path based on a heuristic.

Constraint Satisfaction Problems (CSP): These are problems where:

• Variables need to be assigned values.

• Domains define the possible values each variable can take.

• Constraints are rules that must be satisfied.

Example: Map Coloring:

• Variables = regions

• Domains = {red, green, blue}

• Constraints = No adjacent regions can have the same color
Solving Techniques:

• Backtracking: Tries all combinations recursively.

• Forward Checking: Prevents invalid assignments early.

• Constraint Propagation: Narrows down options using constraints (like Sudoku

logic).
 Module 2: Knowledge Representation and Reasoning

1. Compare and contrast different types of knowledge representation techniques.

Discuss the advantages and disadvantages of each.

o Answer: Knowledge representation techniques include:

▪ Propositional Logic: Uses symbols to represent facts and logical

connectives to form sentences.

▪ Advantages: Simple, well-defined semantics.

▪ Disadvantages: Limited expressiveness, cannot represent

objects or relations directly.

▪ First-Order Logic: Extends propositional logic with objects, relations,

and quantifiers.

▪ Advantages: More expressive, can represent complex

relationships.

▪ Disadvantages: More complex syntax and semantics, inference

can be more challenging.

▪ Semantic Networks: Represent knowledge as a graph of nodes and

edges, showing relationships between concepts.

▪ Advantages: Intuitive representation, good for representing

hierarchical knowledge.

▪ Disadvantages: Can be ambiguous, logic is not always

explicit.
▪ Frames: Structured representations that organize knowledge into slots
and fillers, representing objects or concepts.
▪ Advantages: Good for representing structured objects, can
include default values.

▪ Disadvantages: Can be difficult to represent complex

relationships, logic is limited.

▪ Ontologies: Formal representations of knowledge as a set of concepts

within a domain and the relationships between those concepts.

▪ Advantages: Facilitate knowledge sharing and reuse, support

semantic web technologies.
▪ Disadvantages: Can be complex to design and maintain.
 o The choice of technique depends on the specific requirements of the AI system
and the type of knowledge being represented.

2. Explain the syntax and semantics of propositional logic. Provide examples of how
to represent simple facts and rules using propositional logic.

o Answer:

▪ Syntax: Propositional logic consists of atomic propositions (symbols

representing facts) and logical connectives (AND, OR, NOT,
IMPLIES, EQUIVALENT). Sentences are formed by combining
propositions and connectives.
▪ Semantics: Semantics define the meaning of sentences, determining
their truth value (true or false) based on the truth values of the
propositions and the meaning of the connectives. Truth tables are used
to define the semantics of connectives.

▪ Examples:

▪ Fact: "The sky is blue." represented as SkyBlue

▪ Rule: "If it is raining, then the ground is wet." represented as

Raining IMPLIES GroundWet

3. Describe first-order logic and its components (objects, relations, functions,

quantifiers). How does it differ from propositional logic, and why is it more
expressive?

o Answer:

▪ First-order logic includes:

▪ Objects: Represent entities in the world (e.g., John, a table).

▪ Relations: Represent relationships between objects (e.g.,

Likes(John, Mary)).

▪ Functions: Map objects to other objects (e.g., Father(John) =

Henry).

▪ Quantifiers: Express properties of collections of objects:

▪ Universal Quantifier (∀): "For all" (e.g., ∀x Human(x)

IMPLIES Mortal(x)).

▪ Existential Quantifier (∃): "There exists" (e.g., ∃x

Cat(x) AND Owns(John, x)).

▪ Difference from Propositional Logic: First-order logic can represent

objects, relations between them, and generalize over objects using
 quantifiers, making it much more expressive than propositional logic,
which can only represent simple facts.

4. What are semantic networks and frames? How are they used to represent
relationships and structured knowledge?

o Answer:

▪ Semantic Networks:

▪ Represent knowledge as a graph.

▪ Nodes represent concepts or objects.

▪ Edges represent relationships between them (e.g., "is-a," "has-

a").

▪ Example: A semantic network can represent that "A bird is-a

animal" and "A bird has-a wings."

▪ Frames:

▪ Represent knowledge as data structures with slots and fillers.

▪ Slots represent attributes or properties of an object or concept.

▪ Fillers represent values for those attributes.

▪ Example: A "Bird" frame might have slots for "species,"

"color," "habitat," and "wingspan."

o Both are used to organize and represent knowledge in a structured way, but
semantic networks focus on relationships, while frames focus on the attributes
of objects.
5. Define ontologies and discuss their applications in AI. Explain how ontologies
can be used to improve knowledge sharing and reasoning.

o Answer:
▪ Ontology Definition: An ontology is a formal, explicit specification of
a shared conceptualization. It defines the concepts in a domain and the
relationships between them.

▪ Applications in AI:

▪ Knowledge management
▪ Information integration

▪ Semantic web
▪ Expert systems
 ▪ Improve Knowledge Sharing and Reasoning: Ontologies provide a
common vocabulary and structure, enabling different systems to share
and understand information. They also support reasoning by making
relationships between concepts explicit.
6. Differentiate between deductive and inductive reasoning. Provide examples of
each type of reasoning and discuss their role in AI systems.

o Answer:
▪ Deductive Reasoning:

▪ Derives conclusions that must be true if the premises are true.

▪ Moves from general to specific.

▪ Example: Premise 1: All men are mortal. Premise 2: Socrates is

a man. Conclusion: Socrates is mortal.

▪ Role in AI: Used in logic-based systems, theorem proving, and

verifying the correctness of programs.

▪ Inductive Reasoning:

▪ Draws general conclusions from specific observations.

▪ Conclusions are probable, but not guaranteed.

▪ Moves from specific to general.

▪ Example: Observation: Every swan I have seen is white.

Conclusion: All swans are white.

▪ Role in AI: Used in machine learning, pattern recognition, and

scientific discovery.

7. Explain the concept of rule-based systems. How do they work, and what are their
limitations? Discuss the idea of non-monotone reasoning and its importance.

o Answer:

▪ Rule-Based Systems:
▪ Use a set of "if-then" rules to represent knowledge and make
inferences.

▪ Work by matching the conditions in the "if" part of the rules

with the current situation and executing the "then" part (actions
or conclusions).

▪ Limitations: Can be difficult to handle uncertainty, maintain

consistency in large systems, and learn from data.
 ▪ Non-Monotone Reasoning:

▪ Reasoning where adding new information can invalidate

previous conclusions.

▪ Important for dealing with incomplete or changing information,

as in many real-world situations.

8. Describe probabilistic reasoning and Bayesian networks. How do Bayesian

networks represent uncertain knowledge and support probabilistic inference?

o Answer:

▪ Probabilistic Reasoning:

▪ Deals with uncertainty by using probabilities to represent the

likelihood of events.

▪ Bayesian Networks:

▪ Graphical models that represent probabilistic relationships

between variables.

▪ Nodes represent variables, and edges represent dependencies.

▪ Support probabilistic inference by allowing us to calculate the

probability of some variables given evidence about others.

▪ Represent uncertain knowledge by encoding conditional

probabilities between variables.
 Module 3: Machine Learning
1. Provide an introduction to machine learning. Explain the key concepts and the
importance of machine learning in AI.

o Answer:

▪ Introduction: Machine learning is a field of AI that enables systems to

learn from data without being explicitly programmed.

▪ Key Concepts:

▪ Data: The input used for learning.

▪ Features: The attributes used to describe the data.

▪ Model: The representation of the learned relationship.

▪ Training: The process of learning the model from data.

▪ Prediction: Using the model to make predictions on new data.

▪ Importance: Machine learning allows AI systems to adapt to new
situations, automate decision-making, and extract valuable insights
from large datasets.

2. Describe and compare supervised, unsupervised, and reinforcement learning.

Give examples of tasks that are suitable for each type of learning.

o Answer:

▪ Supervised Learning:
▪ Learns from labeled data (input-output pairs).

▪ Tasks: Classification (predicting categories), Regression

(predicting numerical values).
▪ Examples: Spam detection (classification), predicting house
prices (regression).
▪ Unsupervised Learning:

▪ Learns from unlabeled data.

▪ Tasks: Clustering (grouping similar data points),

Dimensionality reduction (reducing the number of variables).

▪ Examples: Customer segmentation, anomaly detection.

▪ Reinforcement Learning:
▪ Learns through interaction with an environment to maximize a
reward signal.
 ▪ Tasks: Decision-making, control.

▪ Examples: Game playing, robotics control.

3. Explain the decision tree algorithm. How do decision trees work for classification
and regression? Discuss the advantages and disadvantages of decision trees.

o Answer:

▪ Decision Tree Algorithm:

▪ A tree-like structure where each node represents a feature, each

branch represents a decision, and each leaf node represents an
outcome.

▪ Learns by recursively splitting the data based on feature values.

▪ Classification: Predicts categories by following the tree to a leaf node.

▪ Regression: Predicts numerical values, often by averaging the values

of the training data in a leaf node.

▪ Advantages: Easy to understand and interpret, can handle both

categorical and numerical data.

▪ Disadvantages: Can be prone to overfitting, can be unstable (small

changes in data can lead to large changes in the tree).

4. Discuss support vector machines (SVMs). How do SVMs work, and what are
they used for?

o Answer:

▪ Support Vector Machines (SVMs):

▪ Supervised learning algorithm that finds the optimal hyperplane
to separate data into different classes.

▪ Uses kernel functions to handle non-linear data.

▪ How SVMs Work:

▪ Find the hyperplane that maximizes the margin (distance)

between the classes.

▪ Support vectors are the data points closest to the hyperplane.

▪ Uses: Classification and regression.

5. Explain the basics of neural networks. How do neural networks learn, and what
are their applications in AI?
o Answer:
 ▪ Neural Networks:

▪ Inspired by the structure of the brain, composed of

interconnected nodes (neurons) organized in layers.

▪ Learn by adjusting the weights of the connections between

neurons.

▪ How Neural Networks Learn:

▪ Use algorithms like backpropagation to update weights based

on the difference between predicted and actual outputs.

▪ Applications:

▪ Image recognition

▪ Natural language processing

▪ Speech recognition
6. Describe common evaluation metrics for machine learning models. How do we
assess the performance of classification and regression models?
o Answer:

▪ Classification Metrics:

▪ Accuracy: The proportion of correctly classified instances.

▪ Precision: The proportion of correctly predicted positive

instances out of all instances predicted as positive.

▪ Recall: The proportion of correctly predicted positive instances

out of all actual positive instances.

▪ F1-score: The harmonic mean of precision and recall.

▪ Confusion Matrix: A table showing the number of correct and

incorrect predictions for each class.

▪ Regression Metrics:

▪ Mean Squared Error (MSE): The average of the squared

differences between predicted and actual values.

▪ Root Mean Squared Error (RMSE): The square root of the

MSE.

▪ R-squared: A measure of how well the model fits the data.

7. Discuss practical applications of machine learning in AI systems. Provide

examples of how machine learning is used to solve real-world problems.
 o Answer:

▪ Applications:

▪ Recommendation systems: Suggesting products or content to

users.

▪ Fraud detection: Identifying fraudulent transactions.

▪ Medical diagnosis: Assisting in the diagnosis of diseases.

▪ Autonomous vehicles: Enabling self-driving cars.

▪ Natural language processing: Machine translation, sentiment
analysis.
8. Explain the concept of "transfer learning" and its benefits.

o Answer:

▪ Transfer Learning:

▪ A technique where a model trained on one task is reused as the

starting point for a model on a second task.

▪ Benefits:

▪ Can improve learning speed and performance, especially when

the second task has limited data.

▪ Allows leveraging knowledge learned from large datasets.

 Module 4: Natural Language Processing and Robotics

1. What are the basics of natural language processing (NLP)? Describe the
challenges involved in processing human language.

o Answer:

▪ Basics of NLP:

▪ A field of AI that deals with the interactions between computers

and human language.

▪ Involves tasks like understanding, interpreting, and generating

human language.

▪ Challenges:

▪ Ambiguity: Words and sentences can have multiple meanings.

▪ Variability: Language can be expressed in many different
ways.

▪ Context dependence: Meaning can depend on the surrounding

text or situation.

▪ Idioms and figurative language: Non-literal use of language.

2. Explain text processing techniques. How is text processed and prepared for NLP
tasks?

o Answer:

▪ Text Processing Techniques:

▪ Tokenization: Splitting text into individual words or units

(tokens).

▪ Stemming/Lemmatization: Reducing words to their root

form.

▪ Stop word removal: Removing common words that have little

meaning (e.g., "the," "is," "a").

▪ Part-of-speech tagging: Identifying the grammatical role of

each word (e.g., noun, verb, adjective).

3. Discuss language models. How do language models work, and what are they used
for?
o Answer:
 ▪ Language Models:

▪ Statistical models that predict the probability of a sequence of

words.

▪ How Language Models Work:

▪ Learn from large text corpora to estimate the likelihood of word

sequences.

▪ Uses:

▪ Speech recognition

▪ Machine translation
▪ Text generation

▪ Spell checking

4. Describe sentiment analysis and language generation. What are the applications
of these NLP tasks?

o Answer:

▪ Sentiment Analysis:

▪ Identifying the emotional tone or attitude expressed in text

(e.g., positive, negative, neutral).

▪ Applications: Customer feedback analysis, social media

monitoring.

▪ Language Generation:

▪ Generating human-readable text.

▪ Applications: Chatbots, text summarization, content creation.

5. Provide an overview of robotics fundamentals. What are the key components of a

robot?

o Answer:

▪ Robotics Fundamentals:
▪ The field of AI that deals with the design, construction,
operation, and application of robots.

▪ Key Components of a Robot:

▪ Sensors: Gather information about the environment.
▪ Actuators: Motors or other devices that allow the robot to
move.
 ▪ Controller: The "brain" of the robot, processes sensor data and
controls actuators.

▪ Power source: Provides energy to the robot.

6. Explain robot sensor technologies. What types of sensors are used in robotics,
and how do they work?

o Answer:

▪ Robot Sensor Technologies:

▪ Types of Sensors:

▪ Cameras: Provide visual information.

▪ Lidar: Uses lasers to measure distances.

▪ Sonar: Uses sound waves to measure distances.

▪ Touch sensors: Detect contact.

▪ Inertial Measurement Units (IMUs): Measure

orientation and acceleration.

▪ How They Work: Each sensor type uses different physical

principles to convert environmental stimuli into electrical
signals that the robot's controller can process.

7. Discuss robot kinematics and control. How do robots move and interact with
their environment?
o Answer:
▪ Robot Kinematics:

▪ Deals with the motion of robots without considering the forces

that cause the motion.

▪ Forward kinematics calculates the position of the robot's end-

effector given the joint angles.

▪ Inverse kinematics calculates the joint angles required to reach

a desired end-effector position.

▪ Robot Control:

▪ Algorithms and techniques used to control the robot's

movements.

▪ Includes techniques like PID control, motion planning, and

trajectory control.
 8. Describe applications of AI in robotics. Provide examples of how AI is used to
enhance robot capabilities.

o Answer:

▪ Applications of AI in Robotics:

▪ Autonomous navigation: Robots that can move around

without human guidance.

1. Define Natural Language Processing (NLP). What are the main stages involved in
NLP?

Definition:

Natural Language Processing (NLP) is a field of Artificial Intelligence (AI) that focuses on
enabling computers to understand, interpret, and generate human (natural) language. It allows
machines to derive meaning from text or speech in a way that is similar to how humans
process language.
Main Stages of NLP:

1. Lexical Analysis:

o Breaks down text into tokens (words, phrases).

o Also involves morphological analysis, which identifies the structure of words

(e.g., root words, prefixes, suffixes).

2. Syntactic Analysis:

o Also called parsing.

o Analyzes the grammatical structure of a sentence, checking whether the

sequence of words follows grammar rules.

3. Semantic Analysis:

o Assigns meaning to individual words and interprets how they combine to form
meaningful sentences.

o Deals with word sense disambiguation (e.g., “bank” as a financial institution

vs. riverbank).

4. Discourse Integration:

o Ensures that meaning is interpreted in context by considering previous

sentences or conversations.

o For example, interpreting pronouns or repeated references.

5. Pragmatic Analysis:
 o Understands the intended meaning of the language, especially when the
literal meaning doesn’t match the speaker’s intention.

o Example: “Can you pass the salt?” is interpreted as a request, not a question of
ability.

2. Explain Text Processing in NLP. What are the major steps?

Definition:

Text processing is the preparation of raw textual data into a format suitable for machine
learning or NLP algorithms.

Major Steps:

1. Tokenization:

o Splits text into individual units—words or sentences.

o Example: “I love NLP” → [“I”, “love”, “NLP”]

2. Lowercasing:

o Converts all characters to lowercase to ensure consistency.

o Example: “NLP” → “nlp”

3. Stop-word Removal:

o Removes common words that add little meaning (e.g., “is”, “the”, “and”).

4. Stemming and Lemmatization:

o Stemming: Reduces words to their base form by removing suffixes.
(“running” → “run”)
o Lemmatization: Reduces to root form using vocabulary and grammar.
(“better” → “good”)

5. POS Tagging (Part-of-Speech Tagging):

o Assigns grammatical tags to words (e.g., noun, verb).

o Useful for understanding sentence structure.

3. Describe Language Models. What are their types and uses?

Definition:

Language Models (LMs) predict the likelihood of a word sequence. They are essential in
various NLP applications such as speech recognition, machine translation, and chatbots.
Types of Language Models:

1. Statistical Models:

o Based on probabilities.

o N-gram models predict the next word based on the previous (n-1) words.
o Example: Bigram: P(word | previous word)

2. Neural Language Models:

o Use neural networks (e.g., RNN, LSTM).

o Capture better context and semantics than statistical models.

3. Transformer-based Models:

o Advanced models like BERT, GPT, use attention mechanisms.

o Capable of understanding long-range dependencies and context.

Uses:
• Text prediction and generation

• Autocomplete features

• Spell correction

• Machine translation

• Chatbots and virtual assistants

4. What is Sentiment Analysis? Explain how it works and its applications.

Definition:

Sentiment Analysis is the process of identifying and categorizing emotions or opinions in

text, typically classifying them as positive, negative, or neutral.

How It Works:

• NLP tools extract features such as keywords, phrases, or emoticons.

• A classifier (machine learning or deep learning model) evaluates these features.

• It may also use sentiment lexicons (predefined word lists with sentiment scores).

Applications:

• Analyzing product reviews for customer satisfaction

• Social media monitoring for brand feedback
 • Political opinion mining

• Customer support automation

5. What is Natural Language Generation (NLG)? How does it differ from NLP?
Definition:

Natural Language Generation (NLG) is the process of creating human-like text from
structured data (e.g., numbers, tables, databases).
Difference from NLP:

• NLP (Natural Language Processing): Understanding human language.

• NLG (Natural Language Generation): Producing human language.

Applications:

• Automated report writing (e.g., weather, sports, financial)

• Chatbots

• Email or notification generation

• Data storytelling in dashboards

6. Define Robotics. What are the fundamentals of robotics?

Definition:

Robotics is a field that involves the design, construction, and operation of robots—
automated machines that can perform tasks.

Fundamentals:

1. Mechanics: Structure, movement, and body of the robot.

2. Electronics: Sensors and actuators (e.g., motors, LEDs).

3. Control Systems: Logic, programming, and decision-making.

4. Sensors: Enable robots to detect surroundings (e.g., distance, light, sound).

5. Kinematics: Motion planning, joint movement, and positioning.

7. Explain different types of sensors used in robotics.

Types:
1. Proximity Sensors:
 o Detect nearby objects.

o Types: IR sensors, capacitive, inductive.

2. Ultrasonic Sensors:

o Measure distance using sound waves.

o Often used for obstacle avoidance.

3. Infrared (IR) Sensors:

o Detect heat and motion.

o Used in line-following robots or night vision.

4. Gyroscope:

o Measures orientation and rotation.

5. Vision Sensors:
o Capture images or videos.
o Used for object recognition, navigation, and SLAM.

8. What is Robot Kinematics? Differentiate between forward and inverse kinematics.

Definition:

Robot kinematics deals with the motion of a robot without considering forces. It's used for
determining position and orientation.

Types:

• Forward Kinematics:

o Calculates the robot’s end position based on joint parameters (angles,

lengths).

o Easy and deterministic.

• Inverse Kinematics:
o Calculates the joint angles needed to reach a desired position.

o More complex; can have multiple or no solutions.

9. Discuss AI applications in Robotics.

Applications:
1. Autonomous Vehicles:
 o Self-driving cars use AI to navigate roads and traffic.

2. Healthcare Robots:

o Assist in surgeries, patient care, and diagnostics.

3. Industrial Automation:
o Used in assembly lines for repetitive tasks.

4. Humanoid Robots:

o Designed to mimic human behavior for research or assistance.

5. Agricultural Robotics:

o Crop monitoring, harvesting, pesticide spraying.

MODULE 5: Ethical and Societal Implications of AI

1. Key Ethical Considerations:
• Privacy: Protecting user data.

• Bias: Avoiding discrimination based on race, gender, etc.

• Transparency: Making AI decisions understandable.

• Accountability: Assigning responsibility for AI outcomes.

• Job Displacement: Managing the replacement of human labor.

2. AI and Job Displacement:

AI can automate jobs, especially repetitive ones. This may cause unemployment in certain
sectors.
Strategies:

• Reskilling programs

• Entrepreneurship support

• Government subsidies

• Promote collaboration between humans and AI

3. Privacy Concerns in AI Systems:

• AI systems may collect and misuse data.
 • Risks: Data leaks, surveillance, lack of consent.

Solutions:

• Data encryption

• Anonymization
• Clear consent forms

• Legal compliance (e.g., GDPR)

4. How Bias Occurs in AI Algorithms:

Causes:

• Biased training data

• Lack of diversity in datasets

Examples:
• Facial recognition misidentifying certain races

• Gender bias in hiring algorithms

Solutions:

• Train with inclusive data

• Regular audits

• Ethical AI design principles

5. Accountability and Transparency:

• Accountability: Ensuring someone is held responsible for AI errors.

• Transparency: Making AI decisions traceable and understandable.

Importance:

• Builds trust

• Necessary in sensitive domains like finance and healthcare

6. Government and Policy Role:

Governments should:
• Set AI ethics guidelines
 • Promote responsible AI use

• Enforce anti-discrimination laws

• Fund ethical AI research

7. Fairness in AI Decision-Making:

Goal: AI should make unbiased and just decisions for all users.

Approaches:

• Use fairness metrics in algorithms

• Monitor for discrimination

• Perform social impact assessments

8. Societal Implications of AI:

AI can reshape:

• Job markets

• Education systems

• Privacy laws

• Social behaviors

Thus, societal needs and values should be at the core of AI development.