Fall 2025

BECE309L – AI & ML
Knowledge
Representation
Module – 3

Dr. Nitish Katal

 Outline
▪ Knowledge-based agents

▪ Agents based on Propositional Logic

▪ First-order logic
 Terminologies
Knowledge base (KB):

• Key component of a knowledge-based agent.

• These deal with real facts of world.
• It is a mixture of sentences which are explained in knowledge representation language.

Inference Engine(IE):

• It is knowledge-based system engine used to infer new knowledge in the system.

Knowledge Representation (KR):

• The field in AI that focuses on how to formally represent and manage knowledge about the
world in a way that allows computers to use it effectively for reasoning and decision-making.

Ontologies:

• Formal representations that define a set of concepts and the relationships between them
within a domain, providing a shared vocabulary and a structure for reasoning.
 Terminologies
Knowledge Representation Structures
Knowledge Base (KB):

• A collection of knowledge representations that an AI system uses to perform

reasoning tasks.

Taxonomy:

• A hierarchical classification of concepts or entities, often used within

ontologies.

Frames:

• Data structures that represent stereotypical situations or objects by

including attributes and values.
• Each frame is like an object with specific properties.
 Terminologies
Inference and Reasoning:
▪ Inference:
▪ The process of deriving new information or conclusions from existing knowledge.

▪ Reasoning
▪ Deductive Reasoning:
▪ Drawing specific conclusions from general principles or premises.
▪ Inductive Reasoning:
▪ Generalizing from specific instances to broader principles.
▪ Abductive Reasoning:
▪ Inferring the best explanation for a set of observations or facts.
 Terminologies
Examples : Inference and Reasoning:
▪ Deductive Reasoning ▪ Inductive Reasoning ▪ Abductive Reasoning
▪ Premise 1: All mammals have a ▪ Example: You observe that: ▪ Example: Observation: The
backbone. The sun has risen in the east grass is wet in the morning.
▪ Premise 2: A whale is a every morning so far. ▪ Possible Explanation 1: It
mammal. ▪ It rose in the east yesterday. rained during the night.
▪ Conclusion: Therefore, a whale ▪ It rose in the east today. ▪ Possible Explanation 2: The
has a backbone. ▪ Conclusion: Therefore, the sun sprinkler was left on overnight.
▪ In this example, the conclusion will likely rise in the east ▪ Conclusion: Since there are no
is a specific fact drawn from tomorrow. signs of rainfall and the
the general principle that all ▪ Here, the conclusion sprinkler was indeed on, the
mammals have a backbone. generalizes a pattern based on best explanation is that the
specific observations. sprinkler caused the wet grass.
▪ Abductive reasoning involves
inferring the most likely
explanation from the available
evidence.
 Terminologies
▪ Types of Knowledge
▪ Declarative Knowledge:
▪ Knowledge about facts and information
▪ e.g., "The Eiffel Tower is in Paris".
▪ Procedural Knowledge:
▪ Knowledge about how to perform tasks or procedures
▪ e.g., "How to change a tire".
▪ Representation Models
▪ Propositional Logic:
▪ A form of logic where statements are either true or false, and
▪ Logical operations are performed on these statements.
▪ Predicate Logic:
▪ An extension of propositional logic that includes predicates (functions that return true or false) and
quantifiers (e.g., "For all x" or "There exists an x").
 Terminologies
▪ Semantic Networks:
▪ Graph-based structures where nodes represent entities or concepts and edges represent
relationships between them.
▪ Semantic relationships
▪ Hierarchy:
▪ A relationship where one concept is a more general or more specific instance of another (e.g., "Dog"
is a more specific instance of "Animal").
▪ Association:
▪ A relationship where concepts are related but not hierarchically (e.g., "Doctor" is associated with
"Hospital").
 Terminologies
▪ Uncertainty and Ambiguity
▪ Probabilistic Reasoning:
▪ A method of dealing with uncertainty by using probability theory to represent and infer the
likelihood of various outcomes.
▪ Example:
▪ Imagine you're a doctor diagnosing a patient with symptoms of fever, cough, and fatigue. You know
that there are several possible causes, such as the flu, COVID-19, or a common cold. Using
probabilistic reasoning, you assign probabilities based on the prevalence of these illnesses in the
community:
▪ Probability of the flu: 50%
▪ Probability of COVID-19: 30%
▪ Probability of the common cold: 20%
▪ Given the symptoms and their likelihood, you might infer that the flu is the most likely cause.
 Terminologies
▪ Knowledge Acquisition and Learning
▪ Knowledge Acquisition:
▪ The process of gathering and structuring knowledge from various sources to populate a knowledge
base.
▪ Example:
▪ A software company is developing an expert system to assist with medical diagnoses. To populate
the system with knowledge, they interview several doctors and gather information from medical
textbooks and research papers.
▪ They extract rules like:
▪ "If the patient has a high fever and cough, consider testing for the flu.“
▪ "If the patient has loss of taste and smell, test for COVID-19.“
▪ This gathered and structured knowledge is then used to create a knowledge base that the expert
system relies on to make informed decisions.
 Agents that know things
▪ Agents acquire knowledge through perception, learning, language
▪ Knowledge of the effects of actions (“transition model”)
▪ Knowledge of how the world affects sensors (“sensor model”)
▪ Knowledge of the current state of the world
▪ Can keep track of a partially observable world
▪ Can formulate plans to achieve goals
 Knowledge
▪ Knowledge base = set of sentences in a formal language
▪ Declarative approach to building an agent (or other system):
▪ Tell it what it needs to know (or have it Learn the knowledge)
▪ Then it can Ask itself what to do—answers should follow from the KB
▪ Agents can be viewed at the knowledge level
i.e., what they know, regardless of how implemented
▪ A single inference algorithm can answer any answerable question
Knowledge base Domain-specific facts

Inference engine Generic code

 Knowledge based Agents
▪ KBAs are artificial intelligence systems that use knowledge to perform their tasks.
▪ For example, they make deductions, decisions, and conclusions based on available
knowledge.
 Knowledge based Agents
▪ Architecture
▪ KBA receives input from the environment through perception, which the inference
engine processes.
▪ The inference engine communicates with the knowledge base (KB) to determine the
appropriate action based on the knowledge stored in the KB.
▪ The learning element of the KBA regularly updates the KB by incorporating new
knowledge.
 Knowledge based Agents
▪ KBA’s are composed of 2 main modules:
▪ Knowledge Base
▪ Inference Engine
Knowledge Base {KB}
❑ KB of a KBA is a collection of knowledge that the agent uses to make
decisions.
❑ The knowledge can be explicit, such as rules or facts, or implicit, such
as relationships or patterns.
❑ KB is stored in a database or a knowledge representation system.
Knowledge based Agents
Inference Engine {IE}
❑ Inference system of a KBA is responsible for using the knowledge in
the knowledge base to make decisions.
❑ IE uses reasoning methods, such as deduction, induction, and
abduction, to infer new knowledge from existing knowledge.
❑ Generating new facts by the inference system allows the agent to
update its knowledge base.
❑ Operates primarily through two rules: forward chaining and
backward chaining.
 Knowledge based Agents
▪ Forward Chaining: ▪ Backward Chaining:
▪ Forward Chaining the Inference Engine goes ▪ In this, the inference system knows the final
through all the facts, conditions and decision or goal.
derivations before deducing the outcome ▪ This system starts from the goal and works
▪ When based on available data a decision is backwards to determine what facts must be
taken then the process is called as Forwarding asserted so that the goal can be achieved,
chaining. ▪ It works from goal(final decision) and reaches
▪ It works from an initial state and reaches to the initial state.
the goal(final decision). ▪ Example:
▪ Example: ▪ B
▪ A ▪ A -> B
▪ A -> B ▪ A
▪ B or
or ▪ He is sweating.
▪ He is running. ▪ If he is running, he sweats.
▪ If he is running, he sweats. ▪ He is running.
▪ He is sweating.
 Knowledge based Agents
▪ KBA receives a percept and produces an action as output.
▪ The agent is equipped with a knowledge base (KB) with background knowledge of
the real world.
▪ Additionally, the agent has a time counter initially set to zero, indicating the time
elapsed during the process.

▪ Each time the function is executed, the agent performs three operations:
▪ Firstly, it reports to the KB what it has perceived.
▪ Secondly, it asks the KB what action it should take.
▪ Thirdly, it reports to the KB which action it has chosen.
 Knowledge based Agents
▪ MAKE-PERCEPT-SENTENCE
▪ Generates a sentence that indicates that the agent perceived the given percept at
the given time.

▪ MAKE-ACTION-QUERY
▪ Generates a sentence that asks which action should be taken at the current time.

▪ MAKE-ACTION-SENTENCE
▪ Generates a sentence that asserts that the chosen action has been executed.
 Knowledge based Agents
▪ KBA must able to do the following:
▪ An agent should be able to represent states, actions, etc.
▪ An agent should be able to incorporate new percepts
▪ An agent can update the internal representation of the world
▪ An agent can deduce the internal representation of the world
▪ An agent can deduce appropriate actions.
 Levels of KBAs
▪ Knowledge-based agents can be classified into three levels:
▪ Knowledge Level
▪ Knowledge level is the highest level of abstraction in a knowledge-based agent.
▪ It describes what the agent knows and how it uses it to perform tasks.
▪ Knowledge level concerns the representation and organization of knowledge rather than
the implementation details.
▪ Example, suppose an automated taxi agent needs to go from a station A to station B, and it
knows the way from A to B, so this comes at the knowledge level.

▪ Logical Level
▪ Logical level is the intermediate level of abstraction in a knowledge-based agent.
▪ It describes how knowledge is represented and manipulated by inference engine.
▪ Logical story concerns the formal logic used to represent knowledge and make inferences.
▪ Example: At the logical level we can expect to the automated taxi agent to reach to the
destination B.
 Levels of KBAs
▪ Knowledge-based agents can be classified into three levels:
▪ Implementation Level
▪ Implementation level is the lowest level of abstraction in a knowledge-based agent.
▪ It describes how the knowledge and inference engine is implemented using a
programming language.
▪ The implementation level is concerned with the details of the programming
language and the algorithms used to implement the knowledge and inference
engine.
 Actions Performed by the Agent
▪ Inference System is used when we want to update some information
(sentences) in Knowledge-Based System and to know the already
present information.
▪ Mechanism is done by TELL and ASK operations.
▪ Inference include i.e. Producing new sentences from old.
▪ Inference must accept needs when one asks a question to KB and answer should
follow from what has been Told to KB.
▪ Agent also has a KB, which initially has some background Knowledge.
▪ Whenever, agent program is called, it performs some actions
 Actions Performed by the Agent
▪ KBA engage in 3 primary operations to demonstrate intelligent behavior:
▪ TELL:
▪ Agent informs the knowledge base about the information it has perceived
from the environment.
▪ This operation allows the knowledge base to be continually updated with
new facts, ensuring the agent's decisions are based on the most current
information.
▪ ASK:
▪ Agent queries the knowledge base to determine the best course of action
based on the available knowledge.
▪ This operation is crucial for decision-making, allowing the agent to
evaluate various options before taking action.
 Actions Performed by the Agent
▪ PERFORM:
▪ Based on the knowledge base's recommendation, the agent executes
the selected action.
▪ This operation demonstrates the agent's ability to interact with and
impact its environment effectively.
 Designing a KBA
▪ Define the Domain and Scope
▪ Domain Understanding:
▪ Clearly define the domain in which the agent will operate.
▪ Understanding the domain helps in identifying the type of knowledge that needs to be represented and
the complexity of interactions the agent will handle.
▪ Scope Definition:
▪ Determine the scope of the agent's capabilities and functionalities.
▪ This includes specifying the tasks it will perform and the decisions it will make.
▪ Choose the Right Knowledge Representation
▪ Selecting Representation Techniques:
▪ The choice of knowledge representation (KR) technique is critical.
▪ Common KR techniques include semantic networks, frames, rules, and ontologies.
▪ Each has its strengths and is suited to different types of knowledge and reasoning processes.
▪ Representation Language:
▪ Choose a suitable knowledge representation language that can express the complexity of the domain
effectively.
▪ Languages such as OWL (Web Ontology Language), RDF (Resource Description Framework), and rule-based
languages are popular choices.
 Designing a KBA
▪ Develop the Knowledge Base
▪ Gathering Knowledge:
▪ Collect comprehensive and accurate domain knowledge from subject matter experts, literature, and
existing databases.
▪ This knowledge forms the foundation of the agent's decision-making capabilities.
▪ Knowledge Organization:
▪ Organize the knowledge in a structured manner that facilitates efficient retrieval and reasoning.
▪ This includes categorizing knowledge, defining relationships, and establishing hierarchies.

▪ Implement the Inference Engine

▪ Reasoning Mechanisms:
▪ The inference engine should employ reasoning mechanisms suitable for the type of knowledge and the
tasks at hand.
▪ Deductive reasoning, inductive reasoning, and abduction are common approaches.
▪ Algorithm Selection:
▪ Choose algorithms that optimize the reasoning process, considering factors such as the complexity of
queries, the size of the knowledge base, and the need for real-time responses.
 Designing a KBA
▪ Ensure Adaptability and Learning
▪ Incorporating Learning:
▪ Design the agent with mechanisms to learn from new information and experiences.
▪ Techniques such as machine learning algorithms can be integrated to update the knowledge base dynamically.
▪ Feedback Loops:
▪ Implement feedback loops that allow the agent to refine its knowledge and reasoning processes based on outcomes and
external feedback.
▪ Address Ethical and Security Considerations
▪ Ethical Guidelines:
▪ Adhere to ethical guidelines in AI development to ensure that the agent's decisions and actions are fair, transparent, and
respectful of privacy.
▪ Security Measures:
▪ Implement security measures to protect the knowledge base from unauthorized access and ensure the integrity of the
information.
▪ Test and Refine
▪ Prototyping and Testing:
▪ Develop prototypes and conduct thorough testing to evaluate the agent's performance across various scenarios.
▪ This helps in identifying gaps in the knowledge base and inefficiencies in the reasoning process.
▪ Iterative Refinement:
▪ Refine the agent iteratively based on testing outcomes, feedback from users, and evolving domain knowledge.
 Challenges in Designing a KBA
▪ Complexity of Knowledge Representation
▪ Diverse and Complex Domains:
▪ Representing knowledge from complex domains accurately is a significant challenge.
▪ Each domain may have its unique concepts, relationships, and rules that need to be captured in
the knowledge base.
▪ Dynamic Knowledge:
▪ Keeping the knowledge base updated with the latest information and ensuring it adapts to
changes in the domain or environment adds another layer of complexity.

▪ Scalability and Performance

▪ Handling Large Datasets:
▪ As the knowledge base grows, maintaining performance and scalability becomes challenging.
▪ Efficient algorithms and data structures are required to ensure quick access and updates.
▪ Real-Time Processing:
▪ For applications that require real-time decision-making, optimizing the inference engine to
process queries quickly without compromising accuracy is crucial.
 Challenges in Designing a KBA
▪ Ensuring Accuracy and Reliability
▪ Data Quality:
▪ Effectiveness of a knowledge-based agent heavily depends on the quality of the data in its knowledge base.
▪ Ensuring accuracy, relevance, and completeness of this data is a constant challenge.
▪ Error Handling:
▪ Developing robust mechanisms for handling errors or inconsistencies in the knowledge base is essential for
maintaining the reliability of the agent's decisions and actions.

▪ Ethical and Societal Considerations

▪ Bias and Fairness:
▪ Mitigating bias in knowledge-based systems is a significant ethical concern.
▪ Biases in the data or in the decision-making algorithms can lead to unfair or discriminatory outcomes.
▪ Transparency and Explainability:
▪ Ensuring that the agent's decision-making process is transparent and explainable is important for trust and
accountability.
▪ This is especially challenging with complex inference engines and large knowledge bases.
Propositional Logic
 Propositional Logic
▪ A proposition is a declarative statement which is either true or false.
▪ Knowledge representation in logical and mathematical form.
▪ Propositions: Can be either true or false, but it cannot be both.
▪ Propositional logic: Consists of an object, relations or function, and logical connectives.
▪ Connectives: can be said as a logical operator which connects two sentences.
▪ Tautology: Proposition Formula which is always true, and it is also called a valid sentence.
▪ Contradiction: A proposition formula which is always false.
 Propositional Logic
▪ Propositional calculus / Propositional logic : Area of logic which deals with propositions
▪ Propositions are combined together using Logical Connectives or Logical Operators.
▪ Syntax of Propositional Logic:
▪ Defines the allowable sentences for the knowledge representation.
▪ Two types of Propositions:
Atomic Propositions Compound propositions
▪ Atomic propositions are the simple propositions. ▪ Are constructed by combining simpler or
▪ It consists of a single proposition symbol. atomic propositions, using parenthesis and
▪ These are the sentences which must be either logical connectives
true or false.
(a) Atomic Propositions (b) Compound Propositions
 Propositional Logic
▪ Propositional Logic – Logical Connectives
▪ Logical connectives are used to connect two simpler propositions or representing a sentence logically.
▪ A compound propositions can be created with the help of logical connectives.
▪ Five connectives:
▪ Negation:
▪ A sentence such as ¬ P is called negation of P.
▪ A literal can be either Positive literal or negative literal.
▪ Conjunction:
▪ A sentence which has ∧ connective such as, P ∧ Q is called a conjunction.
▪ Example: Roshan is intelligent and hardworking. It can be written as,
▪ P= Roshan is intelligent,
▪ Q= Roshan is hardworking. → P∧ Q.
▪ Disjunction:
▪ A sentence which has ∨ connective, such as P ∨ Q. is called disjunction.
▪ Example: “Ankita is a doctor or Engineer",
▪ Here P= Ankita is Doctor. Q= Ankita is Engineer, so we can write it as P ∨ Q.
 Propositional Logic
▪ Implication:
▪ A sentence such as P → Q, is called an implication.
▪ Implications are also known as if-then rules.
▪ Example: If it is raining, then the street is wet.
▪ Let P= It is raining, and Q= Street is wet, so it is represented as P → Q
▪ Biconditional:
▪ A sentence such as P⇔ Q is a Biconditional sentence,
▪ Example You will pass the exam if and only if you will work hard.
▪ P= You will pass the exam, Q= you will work hard, it can be represented as P ⇔ Q.
Connective Symbols Word Technical Term Example
∧ AND Conjunction 𝑨∧𝑩
∨ OR Disjunction 𝑨∨𝑩
→ Implies Implication 𝑨→𝑩
⇔ If and only if Biconditional 𝑨⇔𝑩
¬ 𝒐𝒓 ~ NOT Negation ¬𝑨 𝒐𝒓 ¬𝑩
 Propositional Logic
▪ Propositional Logic – Truth Table

P Q ¬𝑷 𝑷∧𝑸 𝑷∨𝑸 P→ 𝑸 𝑷⇔𝑸

Negation Conjunction Disjunction Implication Biconditional

True True True True True True

False
True False False True True False

False True False True False False

True
False False False False True True
 Propositional Logic
▪ Propositional Logic – Truth Table

Precedence Operators

1st Precedence Parenthesis In ¬𝑹⋁𝑸 can be

written as ¬𝑹 ⋁𝑸
2nd Precedence Negation

3rd Precedence Conjunction (AND)

4th Precedence Disjunction (OR)

5th Precedence Implication

6th Precedence Biconditional

 Propositional Logic
▪ Properties of operators
 Components of a Propositional Logic Agent
▪ Knowledge Base (KB):
▪ A collection of propositional formulas (or statements) that represent the knowledge the agent has
about the world. These formulas are typically written using logical connectives.
▪ For example,
▪ if the agent knows that "It is raining" (R) and "If it is raining, then the ground is wet" (R → W),
▪ The knowledge base would include these propositions.

▪ Inference Mechanism:
▪ Allows the agent to derive new information from the knowledge base.
▪ Common inference methods include:
▪ Modus Ponens: If we know "P → Q" and "P," we can infer "Q."
▪ Resolution: A method used in automated theorem proving to infer conclusions from the
knowledge base.
▪ Inference mechanism helps the agent make logical conclusions and update its knowledge base
accordingly.
 Components of a Propositional Logic Agent
▪ Query Answering:
▪ When an agent needs to determine if a certain proposition (e.g., "The ground is
wet") is true, it uses the inference mechanism to check if this proposition can be
derived from the knowledge base.
▪ For example,
▪ Given the knowledge base with "R → W" and the fact "R" (It is raining), the
agent can infer "W" (The ground is wet).
 Components of a Propositional Logic Agent
▪ Consider an agent with the following knowledge base:
▪ Propositions:
▪ D: Door is open
▪ M: Motion detected
▪ N: Night time
▪ A: Alarm should sound
▪ Rules:
▪ If the door is open and motion is detected at night, trigger the alarm D ∧ M ∧ N ⇒ A
▪ If motion is detected during the day, don’t trigger the alarm M ∧ ¬N ⇒ ¬A
▪ If door is open at night and there's no motion, don't alarm D ∧ ¬M ∧ N ⇒ ¬A
▪ Facts / Initial Knowledge:
▪ Door is open : D
▪ Motion is detected : M
▪ It is night : N
▪ Inference:
▪ A is true → Alarm must trigger.
 Propositional Logic Agent
▪ Advantages and Limitations
▪ Advantages:
▪ Simplicity: Propositional logic is relatively straightforward, making it easy to
understand and implement.
▪ Decidability: Propositional logic is decidable, meaning there is an algorithm that
can determine whether any given proposition is true or false based on the
knowledge base.
▪ Limitations:
▪ Expressiveness: Propositional logic is limited in its ability to express more complex
statements about the world, especially when dealing with statements involving
variables or quantifiers (which require predicate logic).
▪ Scalability: As the number of propositions grows, the size of the knowledge base
and the complexity of inference can become unwieldy.
 Propositional Logic Agent
▪ Applications
▪ Automated Theorem Proving: Propositional logic is used to prove theorems by checking if they can be
logically derived from a set of axioms.
▪ Planning: In planning problems, agents use propositional logic to generate a sequence of actions that
achieve a desired goal.
▪ Diagnosis: Agents use propositional logic to identify the cause of problems based on symptoms and
known relationships.
▪ Knowledge Representation: Used in expert systems to encode domain-specific knowledge, enabling
systems to make informed decisions or provide recommendations.
▪ Automated Reasoning: Employed in legal reasoning systems to evaluate legal arguments or in
configuration systems to ensure that all parts of a product configuration are compatible.
▪ Game Playing and Strategy: Used in simple board games or puzzle-solving applications where the
environment and rules are well-defined.
▪ Intelligent Tutoring Systems: Used in educational technology to provide personalized feedback and
guidance based on a student’s responses.
Propositional Logic Agent
 Propositional Logic Agent
Smart Home Assistant: Agent controls lights and heating based on user preferences and
sensor data.

❑ KB Example: Let
❑ M: Motion detected in room
❑ D: It is dark
❑ L: Light should be on
❑ H: Heating should be on
❑ C: It is cold
❑ Rules:
❑ (M ∧ D) ⇒ L
❑ C⇒H
❑ Facts:
❑ M is true (motion detected)
❑ D is true (dark)
❑ Inference: ???
 Propositional Logic Agent
Smart Home Assistant: Agent controls lights and heating based on user preferences and
sensor data.

❑ KB Example: Let
❑ M: Motion detected in room
❑ D: It is dark
❑ L: Light should be on
❑ H: Heating should be on
❑ C: It is cold
❑ Rules:
❑ (M ∧ D) ⇒ L
❑ C⇒H
❑ Facts:
❑ M is true (motion detected)
❑ D is true (dark)
❑ Inference:
❑ L is true (light should be on)
 Propositional Logic Agent
Traffic Signal Controller: Control car lights based on rules.

❑ KB Example: Let
❑ R: Red light is on
❑ G: Green light is on
❑ S: Car should stop
❑ M: Car should move
❑ Rules:
❑ R⇒S
❑ G⇒M
❑ ¬(R ∧ G) (cannot be both red and green)

❑ Facts:
❑ R is true
❑ Inference: ???
 Propositional Logic Agent
Traffic Signal Controller: Control car lights based on rules.

❑ KB Example: Let
❑ R: Red light is on
❑ G: Green light is on
❑ S: Car should stop
❑ M: Car should move
❑ Rules:
❑ R⇒S
❑ G ⇒ M¬(R ∧ G) (cannot be both red and green)

❑ Facts:
❑ R is true
❑ Inference:
❑ S is true (car must stop)
 Propositional Logic Agent
Traffic Signal Controller: Control car lights based on rules.

❑ KB Example: Let
❑ R: Red light is on
❑ G: Green light is on
❑ S: Car should stop
❑ M: Car should move
❑ Rules:
❑ R⇒S
❑ G ⇒ M¬(R ∧ G) (cannot be both red and green)

❑ Facts:
❑ R is true
❑ Inference:
❑ S is true (car must stop)
 Propositional Logic Agent
A robot is navigating a 2D grid (like a maze or terrain). It must determine which adjacent cells are safe
or contain obstacles, using its local sensor that can detect if an obstacle is nearby
(in any of the 4 directions: N, S, E, W).

Symbol Table
Symbol Meaning
𝐶𝑖𝑗 Cell (i,j) is Clear (no obstacle)
𝑂𝑖𝑗 Cell (i,j) contains an Obstacle
𝑆𝑖𝑗 Sensor at cell (i,j) is triggered (i.e., obstacle nearby)
𝐴𝑖𝑗 Agent is currently at cell (i,j)
¬ Logical NOT
∧,∨, ⇒ Logical AND, OR, and IMPLIES
 Propositional Logic Agent
Rules
Rule 1: Sensor implication (triggered)
If the sensor at (i,j) is triggered → at least one of the adjacent cells has an obstacle.

𝑺𝒊𝒋 ⇒ 𝑶𝒊 𝒋−𝟏 ∨ 𝑶𝒊 𝒋+𝟏 ∨ 𝑶 𝒊−𝟏 𝒋 ∨ 𝑶 𝒊+𝟏 𝒋

Rule 2: No sensor → no adjacent obstacle

If the sensor is not triggered → all adjacent cells are clear (not obstacles).

¬𝑺𝒊𝒋 ⇒ ¬𝑶𝒊 𝒋−𝟏 ∧ ¬𝑶𝒊 𝒋+𝟏 ∧ ¬𝑶 𝒊−𝟏 𝒋 ∧ ¬𝑶 𝒊+𝟏 𝒋

Rule 3: Combining neighbor sensors for deduction

If sensor at (2,3) is triggered, but adjacent sensors like (2,2) and (1,3) are not, then only some cells near
(2,3) can have an obstacle.

𝑆𝑖𝑗 ∧ ¬𝑆 𝑖−1 𝑗 ∧ ¬𝑆 𝑖+1 𝑗 ⇒ 𝑂𝑖 𝑗−1 ∨ 𝑂𝑖 𝑗+1

 Propositional Logic Agent
Facts
Sample KB Facts: The robot has the following sensory inputs:
Agent is at 𝑨𝟐𝟐 ∶ ¬𝑺𝟐𝟐 , 𝑺𝟐𝟑 , ¬𝑺𝟐𝟏 ¬S22 → No sensor triggered at current cell (2,2)
S23 → Sensor at East cell (2,3) is triggered
¬S21 → Sensor at West cell (2,1) is not triggered

So:
At (2,2) → no obstacle near the agent
At (2,3) → obstacle nearby, but not necessarily in (2,3)
At (2,1) → no obstacle near that cell
 Propositional Logic Agent
Facts
Sample KB Facts:
Agent is at 𝑨𝟐𝟐 ∶ ¬𝑺𝟐𝟐 , 𝑺𝟐𝟑 , ¬𝑺𝟐𝟏

Inferences
Apply Rule 2 to (2,2):
From ¬𝑺𝟐𝟐 ¬S22 ⇒ (¬O21 ∧ ¬O23 ∧ ¬O12 ∧ ¬O32)
infer: ¬𝑶𝟐𝟏 ∧ ¬𝑶𝟐𝟑 ∧ ¬𝑶𝟏𝟐 ∧ ¬𝑶𝟑𝟐
Inference 1:
Adjacent cells to (2,2) are all safe.
 Propositional Logic Agent
Facts
Sample KB Facts:
Agent is at 𝑨𝟐𝟐 ∶ ¬𝑺𝟐𝟐 , 𝑺𝟐𝟑 , ¬𝑺𝟐𝟏

Inferences
From ¬𝑺𝟐𝟐
infer: ¬𝑶𝟐𝟏 ∧ ¬𝑶𝟐𝟑 ∧ ¬𝑶𝟏𝟐 ∧ ¬𝑶𝟑𝟐

From ¬𝑺𝟐𝟏 Apply Rule 2 to (2,1):

¬S21 ⇒ (¬O11 ∧ ¬O22 ∧ ¬O31 ∧ ¬O20)
infer: ¬𝑺𝟐𝟏 ⇒ (¬𝑶𝟏𝟏 ∧ ¬𝑶𝟐𝟐 ∧ ¬𝑶𝟑𝟏 ∧ ¬𝑶𝟐𝟎)
Inference 2:
Adjacent cells to (2,1) are all safe.
Propositional Logic Agent
Facts
Sample KB Facts:
Agent is at 𝑨𝟐𝟐 ∶ ¬𝑺𝟐𝟐 , 𝑺𝟐𝟑 , ¬𝑺𝟐𝟏

Inferences
From ¬𝑺𝟐𝟐
infer: ¬𝑶𝟐𝟏 ∧ ¬𝑶𝟐𝟑 ∧ ¬𝑶𝟏𝟐 ∧ ¬𝑶𝟑𝟐

From ¬𝑺𝟐𝟏
infer: ¬𝑺𝟐𝟏 ⇒ (¬𝑶𝟏𝟏 ∧ ¬𝑶𝟐𝟐 ∧ ¬𝑶𝟑𝟏 ∧ ¬𝑶𝟐𝟎) Apply Rule 1 to (2,3):
𝑺𝟐𝟑 ⇒ (𝑶𝟐𝟐 ∨ 𝑶𝟐𝟒 ∨ 𝑶𝟏𝟑 ∨ 𝑶𝟑𝟑)
From 𝑺𝟐𝟑
infer: 𝑺𝟐𝟑 ⇒ (𝑶𝟐𝟒 ∨ 𝑶𝟏𝟑 ∨ 𝑶𝟑𝟑) Inference 3:
But — we already know that:
¬O22 (no obstacle at current cell)
¬O23 (the cell where sensor is triggered itself)
 First-order Logic /
Predicate Logic /
First-order Predicate Logic
 First Order Logic (FOL)
▪ Propositional Logic (PL), which can only represent the facts, which are either true or false.
▪ PL is not sufficient to represent the complex sentences or natural language statements.
▪ PL has very limited expressive power.
▪ Consider the following sentence, which we cannot represent using PL logic.
▪ "Some humans are intelligent", or
▪ "Sachin likes cricket.“

▪ First-order logic is another way of knowledge representation: An extension to PL.

▪ FOL is sufficiently expressive to represent the natural language statements in a
concise way.
▪ FOL is a powerful language that develops information about the objects in a more
easy way and can also express the relationship between those objects.
 First Order Logic : Components
▪ First-order logic (like natural language) does not only assume that the world contains facts like
propositional logic but also assumes the following things in the world:

▪ Objects: A, B, people, numbers, colors, theories, ▪ Variables: Variables are symbols that can
squares, … represent any object in the domain.
▪ Examples: Variables such as x, y, and z can
▪ Relations: It can be unary relation such as: red, represent any object in the domain.
round, is adjacent, or n-any relation such as: the
sister of, brother of, has color, comes between, … ▪ Predicates: Predicates represent properties of
objects or relationships between objects.
▪ Function: Father of, best friend, third inning of, end
of, … ▪ Examples: P(x) could mean “x is a person”,
while Q(x, y) could mean “x is friends with y”.
▪ Constants: Constants are symbols that
represent specific objects in the domain. ▪ Functions: Functions map objects to other
objects.
▪ Examples: If a, b, and c are constants, they
might represent specific individuals like Alice, ▪ Examples: f(x) could represent a function that
Bob, and Charlie. maps an object x to another object, like “the
father of x”.
 First Order Logic : Components
• Logical Connectives:
• These include ∧ (and), ∨ (or), ¬ (not), → (implies), and (if and only if).
• Examples: P(x) ∧ Q(x, y) means “P(x) and Q(x, y) are both true”.
• Equality:
• States that two objects are the same.
• Examples: x = y asserts that x and y refer to the same object.

Syntax of First-Order Logic

The syntax of FOL defines the rules for constructing well-formed formulas:
▪ Atomic Formulas: The simplest formulas, which can be predicates applied to terms
▪ e.g., 𝑃(𝑎), 𝑄(𝑥, 𝑦)
▪ Complex Formulas: Formed by combining atomic formulas using logical
connectives and quantifiers.
▪ e.g., ∀𝒙 (𝑷(𝒙) ∨ ¬𝑸(𝒙, 𝒇(𝒚)))
 First Order Logic : Components
Semantics of First-Order Logic
The semantics define the meaning of FOL statements:
▪ Domain: A non-empty set of objects over which the variables range.

▪ Interpretation: Assigns meanings to the constants, functions, and predicates,

▪ Specifying which objects the constants refer to,
▪ Which function the function symbols denote, and
▪ Which relations the predicate symbols denote.

▪ Truth Assignment: Determines the truth value of each formula based on the
interpretation.
 First Order Logic : Key Components
▪ Two main components:
▪ Syntax:
▪ Syntax represents the rules
to write expressions in First
Order Logic in Artificial
Intelligence.

▪ Semantics:
▪ Semantics refers to the
techniques that we use to
evaluate an expression of
First Order Logic in AI.
▪ These techniques use
various known relations and
facts of the respective
environment to deduce the
boolean value of the given
First Order Logic expression.
 First Order Logic : Key Components
▪ Atomic Sentences:
▪ Atomic sentences are the most basic expressions of First Order Logic in AI.
▪ These sentences comprise a predicate followed by a set of terms inside a
parenthesis.

▪ Formally stating, the structure of an atomic sentence looks like the following.
▪ Predicate 1 ( term 1, term 2, term 3,...)
▪ Predicate 2 ( term 2​, term 4​,...)
 First Order Logic : Key Components
▪ Complex Sentences:
▪ Complex sentences can be constructed by combining atomic sentences using connectives like
▪ AND (∧), OR (∨), NOT (¬), IMPLIES (⇒), IF AND ONLY IF (⇔) etc.

▪ Formally stating, if c1, c2, ... c3​,c4​... represent connectives, a complex sentence in First Order Logic in
AI can be defined as follows.
▪ Predicate 1( term 1, term 2,...) c1 Predicate 2( term 1, term 2,...)

▪ For example, in the expression, Polygon (Rectangle), which is a translation of "Rectangle is a

polygon." in First Order Logic in Artificial Intelligence,
▪ "Rectangle" is the subject, and "Polygon" represents a predicate.
 First Order Logic : Key Components
▪ Quantifiers in First-Order Logic: Quantify any entity in a given environment.
▪ Quantification refers to the identification of the total number of an entity that is present in
the environment and satisfies a given expression in First Order Logic in Artificial Intelligence.

▪ Quantifiers enable us to determine the range and scope of a variable in a logical expression.
▪ Two types of quantifiers :
▪ Universal &
▪ Existential Quantifier
 First Order Logic : Key Components
▪ Universal Quantifier:
▪ Is a symbol in a logical expression that signifies that the given expression is true in its range for all
instances of the concerned entity.
▪ Represented by the symbol ∀ (an inverted A).
▪ If x is a variable, then ∀x is read as "For all x" or "For every x" or "For each x".

▪ For example:
▪ Let us take the sentence, "All cats like fish".
▪ Let us take a variable x which can take the value of "cat".
▪ Let us take a predicate cat (x) which is true if x is a cat.
▪ Similarly, let us take another predicate likes (x, y) which is true if x likes y.
▪ Therefore, using the universal quantifier ∀, we can write
∀𝒙, 𝒄𝒂𝒕 (𝒙) ⟹ 𝒍𝒊𝒌𝒆𝒔 (𝒙, 𝒇𝒊𝒔𝒉 ).
▪ This expression is read as "For all x, if x is a cat, then x likes to fish".
 First Order Logic : Key Components
▪ Existential Quantifier
▪ Is a symbol in a logical expression that signifies that the given expression is true in its range for at
least one of the instances of the concerned entity.
▪ It is represented by the symbol ∃ (an inverted E).
▪ If x is a variable, then ∃x is read as "There exists x" or "For some x" or "For at least one x".

▪ For example: "Some students like ice cream".

▪ Let us take a variable x which can take the value of "student".
▪ Let us take a predicate student (x), which is true if x is a student.

▪ Similarly, let us take another predicate likes (x,y), which is true if xx likes yy.
▪ Therefore, using the existential quantifier ∃, we can write
▪ ∃ 𝒙 𝒔𝒕𝒖𝒅𝒆𝒏𝒕 (𝒙) ∧ 𝒍𝒊𝒌𝒆𝒔 (𝒙, 𝒊𝒄𝒆 − 𝒄𝒓𝒆𝒂𝒎 )
▪ This expression reads, "There exists some x such that x is a student and also likes ice cream".
 First Order Logic : Key Components
▪ Examples of FOL using quantifier:
▪ All birds fly.
In this question the predicate is "fly(bird)."
And since there are all birds who fly so it will be represented as follows.
∀𝒙 𝒃𝒊𝒓𝒅(𝒙) → 𝒇𝒍𝒚(𝒙)

▪ Everyone respects their parents.

In this question, the predicate is "respect(x, y)," where x=man/women, and y= parent.
Since there is every one so will use ∀, and it will be represented as follows:
∀𝒙 𝑬𝒗𝒆𝒓𝒚𝒐𝒏𝒆 (𝒙) → 𝒓𝒆𝒔𝒑𝒆𝒄𝒕𝒔 (𝒙, 𝒑𝒂𝒓𝒆𝒏𝒕)
 First Order Logic : Key Components
▪ Examples of FOL using quantifier:
▪ Some boys play cricket.
In this question, the predicate is "play(x, y)," where x= boys, and y= game. Since there
are some boys so we will use ∃, and it will be represented as:
∃𝒙 𝒃𝒐𝒚𝒔(𝒙) → 𝒑𝒍𝒂𝒚(𝒙, 𝒄𝒓𝒊𝒄𝒌𝒆𝒕)

▪ Not all students like both Mathematics and Science.

In this question, the predicate is "like(x, y)," where x= student, and y= subject.
Since there are not all students, so we will use ∀ with negation, so following
representation for this:
¬∀ (𝒙) [ 𝒔𝒕𝒖𝒅𝒆𝒏𝒕(𝒙) → 𝒍𝒊𝒌𝒆(𝒙, 𝑴𝒂𝒕𝒉𝒆𝒎𝒂𝒕𝒊𝒄𝒔) ∧ 𝒍𝒊𝒌𝒆(𝒙, 𝑺𝒄𝒊𝒆𝒏𝒄𝒆)]
 First Order Logic : Key Components
▪ While using quantifiers in writing expressions for FOL, we need to keep the following points in mind.
▪ Main connective for the universal quantifier ∀ is the implication (⟹).
▪ Main connective for existential quantifier ∃ is and (∧).

▪ Types of Variables: Free and Bound Variables

▪ Based upon their interaction with the quantifiers.

▪ Free Variables: ▪ Bound Variables:

▪ Those variables that do not come under ▪ Those variables that occur inside the
the scope of the quantifier. scope of the quantifier.
▪ For instance, in an ▪ For instance, in an
expression ∀𝒙∃𝒚𝑷(𝒙, 𝒚, 𝒛), z is a free expression ∀𝑥∃𝑦𝑃(𝑥, 𝑦, 𝑧), x and y are
variable because it doesn't come under the bound variables because they occur inside
scope of any quantifier. the scope of the quantifiers.
 First Order Logic : Challenges
▪ Complexity:
▪ Representing certain real-world domains accurately in FOL can lead to complex and unwieldy
formulas, making reasoning and inference computationally expensive.
▪ Expressiveness Limitations:
▪ FOL has limitations in representing uncertainty, vagueness, and probabilistic relationships, which are
common in many AI applications.
▪ Knowledge Acquisition:
▪ Encoding knowledge into FOL requires expertise and manual effort, making it challenging to scale and
maintain large knowledge bases.
▪ Inference Scalability:
▪ Reasoning in FOL can be computationally intensive, especially in large knowledge bases, requiring
efficient inference algorithms and optimization techniques.
▪ Handling Incomplete Information:
▪ FOL struggles with representing and reasoning with incomplete or uncertain information, which is
common in real-world applications.
 First Order Logic : Limitations
▪ Inability to Represent Recursive Structures:
▪ FOL cannot directly represent recursive structures, limiting its ability to model certain types of relationships and
processes.
▪ Lack of Higher-Order Reasoning:
▪ FOL lacks support for higher-order logic, preventing it from representing and reasoning about properties of
predicates or functions.
▪ Difficulty in Representing Context and Dynamics:
▪ FOL struggles with representing dynamic or context-dependent knowledge, such as temporal relationships or
changes over time.
▪ Limited Representation of Non-binary Relations:
▪ FOL primarily deals with binary relations, making it less suitable for representing complex relationships involving
multiple entities.
▪ Difficulty in Handling Non-monotonic Reasoning:
▪ FOL is not well-suited for non-monotonic reasoning, where new information can lead to retraction or
modification of previously inferred conclusions.