Knowledge Representation

Knowledge representation is one of the top priorities in artificial intelligence research. Knowledge is made up of facts, concepts, rules, etc. Knowledge can be presented in different forms. For example, language speaks the mental image of one's thoughts, by writing or can be expressed by imaging or by storing a collection of magnetic stains on a computer. Its use will depend on the type of problem it is being used to solve and the type of estimation method used. For example, a game of cards or cards may be mentioned. Some rules and card values are used in card games. This is the best way to present it. For example, the four types of cards are Spades, Hearts, Diamonds, and Clubs. Who is evaluated and their values are ace, 2,3,....10, Golam, Saheb, Bibi are used as symbolic pairs. (e.g., BB or Queen of Hearts <Queen, Hearts> is called). In playing bridge, both suit and value are important, while in blackjack, only face value is important. The solution can be simplified by state diagram representation. For example, The Tower of Hanoi problem. Intermediate states are used to move from the initial state to the final state to solve using rules and conditions.

Several other types of representation methods have gained popularity among AI researchers. Perhaps the most popular of these are propositional logic and first-order predicate logic. This is important because there are very few methods that have advanced theory, Reasonable expressive ability, and formation of valid decisions.

Other representation methods include frames and associative networks, fuzzy logic, modal logic, Object Oriented Method, etc. Here are explained some important knowledge representation methods briefly.

1. Semantic Networks: Semantic networks represent knowledge as a network of interconnected nodes, where each node represents a concept or an entity, and the edges represent relationships between them. This approach is useful for representing hierarchical relationships and associations between concepts.

Example: Representing relationships between animals.

Nodes: {Cat, Dog, Pet, Mammal}

Edges: {Cat is a Pet, Dog is a Pet, Cat is a Mammal, Dog is a Mammal}

2. Frames: Frames are structures that represent a collection of attributes and values associated with a particular concept or object. Each frame consists of slots, which hold specific information about the concept. Frames help capture the properties, behaviors, and relationships of entities in a domain.

Example: Representing a car frame.

Frame: Car, Slots: {Make: Toyota, Model: Camry, Year: 2023, Color: Blue}

3. Logic-based Representations: Logic-based representations, such as propositional logic and First-Order Logic (FOL) or Predicate Logic, use logical formulas to represent knowledge. They define the relationships between objects, facts, and rules using logical operators like AND, OR, and NOT. These representations enable formal reasoning and inference. We explain here propositional logic and first-order logic.

Prepositional logic, or propositional logic, is a branch of mathematical logic that deals with propositions and their logical relationships using symbolic representation and formal rules. In propositional logic, propositions are statements that can be either true or false. Propositions are represented using propositional variables, such as P, Q, R, etc. These variables can stand for any statement or atomic proposition.

Propositional logic provides a set of logical operators that can be used to combine or manipulate propositions. The commonly used logical operators in propositional logic include:

i Negation (NOT): Denoted by the symbol ¬ or ~, it negates a proposition, flipping its truth value. For example, if P is true, then ¬P is false, and vice versa. Let's consider the proposition P: “It is sunny.” The negation of P, denoted as ¬P or ~P, would be “It is not sunny.”

ii Conjunction (AND): Denoted by the symbol ∧, it represents the logical “and” operation between two propositions. It evaluates to true only when both propositions are true. For example, if P is true and Q is true, then P ∧ Q is true. Otherwise, it is false. Let's consider two propositions:

P: “It is raining.”

Q: “I have an umbrella.” The conjunction of P and Q, denoted as P ∧ Q, would be “It is raining and I have an umbrella.”

iii Disjunction (OR): Denoted by the symbol ∨, it represents the logical “or” operation between two propositions. It evaluates to true if at least one of the propositions is true. For example, if P is true or Q is true (or both), then P ∨ Q is true. It is false only when both propositions are false. Let's consider two propositions:

P: “It is Monday.”

Q: “It is Friday.” The disjunction of P and Q, denoted as P ∨ Q, would be “It is either Monday or Friday.”

iv Implication (IF-THEN): Denoted by the symbol →, it represents the logical “if-then” relationship between two propositions. It states that if the antecedent proposition (left-hand side) is true, then the consequent proposition (right-hand side) must also be true. If the antecedent is false, the implication is true. For example, if P is true and Q is true, then P → Q is true. If P is false, the implication is still true regardless of the truth value of Q. Let's consider two propositions:

P: “It is snowing.”

Q: “I will wear a coat.” The implication of P implies

Q, denoted as P → Q, would be “If it is snowing, then I will wear a coat.”

v Biconditional (IF AND ONLY IF): Denoted by the symbol ↔, it represents the logical “if and only if” relationship between two propositions. It states that the two propositions have the same truth value. If both propositions are true or both are false, then the biconditional is true. For example, if P is true and Q is true, or if P is false and Q is false, then P ↔ Q is true. It is false if the truth values of P and Q differ. Let's consider two propositions:

P: “The car is red.”

Q: “The car is new.” The biconditional of P and Q, denoted as P ↔ Q, would be “The car is red if and only if it is new.”

Using these logical operators, propositions can be combined to form complex expressions, and their truth values can be determined based on the truth values of the atomic propositions and the rules of logical operators.

Propositional logic provides a foundation for reasoning and formalizing logical arguments and deductions. It is widely used in mathematics, computer science, philosophy, and artificial intelligence to represent and analyze logical relationships between statements.

First-Order Logic (FOL), also known as Predicate Logic or First-Order Predicate Calculus, is an extension of propositional logic that allows for a more expressive and precise representation of logical statements involving objects, properties, and relations. It is examples of objects, relations, and functions:
•Objects: people, houses, numbers, theories, Ronald McDonald, colors, baseball games, wars, centuries, etc.
•Relations: these can be unary relations or properties such as red, round, bogus, prime, multistoried, etc, and more general n-ary relations such as brother of, bigger than, inside, part of, has color, occurred after, owns, comes between, etc.
•Functions: father of, best friend, third inning of, one more than, the beginning of, Indeed, almost any assertion can be thought of as referring to objects and properties or relations. Some examples follow:
• “One plus two equals three.”
Objects: one, two, three, one plus two; Relation: equals; Function: plus. (“One plus two” is a name for the object that is obtained by applying the function “plus” to the objects “one” and “two.” “Three” is another name for this object.)

• “Squares neighboring the wumpus are smelly.”

Objects: wumpus, squares;

Property: smelly; Relation: neighboring.

• “Evil King John ruled England in 1200.”
Objects: John, England, 1200;

Relation: ruled;

Properties: evil, king.

In First-Order Logic, we introduce additional elements to capture the structure and semantics of statements. These elements include:

i Quantifiers: FOL introduces quantifiers to express statements about objects in a domain. There are two types of quantifiers:

· Universal Quantifier (∀): Denoted by ∀, it expresses that a statement holds for all objects in the domain. For example, ∀x P(x) means “For all x, P(x) is true.”, ∀x P, where P is any logical expression, says that P is true for every object x. For example “All kings are persons,” is written in first-order logic as ∀ x King(x) ⇒ Person(x).
∀ is usually pronounced “For all . . .”. (Note that the upside-down A stands for “all.”)
Thus, the sentence says, “For all x, if x is a king, then x is a person.” The symbol x is called a variable. By convention, variables are lowercase letters. A variable is a term all by itself and as such can also serve as the argument of a function—for example, LeftLeg(x).

· Existential Quantifier (∃): Denoted by ∃, it expresses that there exists at least one object in the domain for which a statement holds. For example, ∃x P(x) means “There exists an x for which P(x) is true.”

ii Variables: FOL introduces variables to represent unspecified objects in statements. Variables are denoted by symbols such as x, y, z, etc. They can be universally or existentially quantified.

iii Predicates: Predicates are used to express properties or relations between objects. They are often represented by uppercase letters or symbols and take one or more arguments. For example, P(x) can represent a property of an object x, and R(x, y) can represent a relation between objects x and y.

iv Functions: FOL allows the use of functions to describe operations or transformations on objects. Functions take arguments and return a value. For example, f(x) can represent a function that maps an object x to a value.

v Constants: Constants represent specific objects in the domain. They are often represented by lowercase letters or symbols. For example, a, b, and c can be constants representing specific individuals.

Using these elements, First-Order Logic allows us to express complex statements involving quantification, predicates, functions, and variables. It provides formal and precise language to reason about relationships, properties, and structures within a given domain.

Example of first-order logic:

A. Example: Representation of a simple logical statement. Statement: All cats are mammals. FOL Representation:

∀x (Cat(x) → Mammal(x))

This statement can be represented in FOL as “For all x, if x is a cat, then x is a mammal.”

B. Example: Representation of a relationship between objects. Statement: John is the father of Mary. FOL Representation: Father (John, Mary)

This statement represents the relationship between John and Mary using the predicate “Father.”

C. Example: Representation of existential quantification. Statement: There exists a brown dog. FOL Representation: ∃x (Dog(x) ∧ Brown(x))

This statement can be represented in FOL as “There exists an x such that x is a dog and x is brown.”

D. Example: Representation of negation. Statement: It is not true that all birds can fly. FOL Representation:

¬(∀x (Bird(x) → CanFly(x)))

This statement represents the negation of the statement “For all x, if x is a bird, then x can fly.”

E. Example: Representation of logical equivalence. Statement: A and B are equivalent. FOL Representation:

(A ↔ B)

This statement represents the logical equivalence between propositions A and B.

F. Example: Representation of transitive property. Statement: If A is taller than B, and B is taller than C, then A is taller than C. FOL Representation:

∀x, y, z ((Taller(x, y) ∧ Taller(y, z)) → Taller(x, z))

This statement represents the transitive property of “Taller” in FOL.

Some more examples:

G. Marcus tried to assassinate Caesar. Predicate Logic Representation: TryAssassinate(Marcus, Caesar)

H. All Pompeians were Roman. Predicate Logic Representation: ∀x(Pompeian(x) → Roman(x))

I. All Romans were either loyal to Caesar or hated him. Predicate Logic Representation: ∀x(Roman(x) → (LoyalToCaesar(x) ∨ HateCaesar(x)))

J. Everyone is loyal to someone. Predicate Logic Representation: ∀x∃y(LoyalTo(x, y))

K. People only try to assassinate rulers they are not loyal to. Predicate Logic Representation: ∀x∀y((Person(x) ∧ Ruler(y)) → (TryAssassinate(x, y) → ¬LoyalTo(x, y)))

L. vi) Some birds can fly. Predicate Logic Representation: ∃x(Bird(x) ∧ CanFly(x))

M. No humans are immortal. Predicate Logic Representation: ¬∃x(Human(x) ∧ Immortal(x))

N. Every student passed the exam. Predicate Logic Representation: ∀x(Student(x) → PassedExam(x))

O. There exists a square with all sides equal. Predicate Logic Representation: ∃x(Square(x) ∧ ∀y(SideLength(x, y)))

P. All cats chase mice. Predicate Logic Representation: ∀x(Cat(x) → ∃y(Mice(y) ∧ Chase(x, y)))

First-Order Logic is widely used in various fields, including mathematics, computer science, philosophy, and artificial intelligence, to model and reason about the real world, specify rules, constraints, and relationships, and facilitate automated reasoning and inference.

4. Ontologies: Ontologies provide a systematic way of representing knowledge by defining concepts, relationships, and properties in a domain. They often use a hierarchical structure to represent different levels of abstraction and allow for the specification of domain-specific knowledge. Example: Representing a medical domain ontology.

Concepts: {Disease, Symptom, Treatment}

Relationships: {Disease hasSymptom Symptom, Disease requiresTreatment Treatment}

5. Rule-based Systems: Rule-based systems represent knowledge as a set of rules in the form of “if-then” statements. These rules define conditions (antecedents) and corresponding actions (consequents). Rule-based systems are useful for encoding expert knowledge and capturing complex decision-making processes. Example: A rule-based system for traffic light control.

Rules: IF the traffic light is red THEN stop, IF the traffic light is green THEN go, IF the traffic light is yellow THEN prepare to stop.

6. Bayesian Networks: Bayesian networks represent knowledge using probabilistic relationships between variables. They utilize graphical models to represent dependencies and uncertainties, allowing for reasoning under uncertainty and making probabilistic inferences. Example: Predicting the probability of a student passing an exam based on study hours and previous grades.

Variables: {StudyHours, PreviousGrades, PassExam}

Dependencies: P(PassExam | StudyHours, PreviousGrades) = f(P(StudyHours), P(PreviousGrades)).

7. Neural Networks: Neural networks can also be used for knowledge representation, particularly in the context of machine learning. Neural networks learn patterns and representations from data, enabling the capture and utilization of implicit knowledge. Example: Image classification of different types of flowers using a deep convolutional neural network (CNN) trained on a flower dataset. In this example, a neural network, specifically a deep convolutional neural network, is trained on a dataset of labeled images containing different types of flowers. The neural network learns to classify new images into the appropriate flower categories based on the learned patterns and features.



Thank you for reading this post and encourage me to grow this blog further