Knowledge

Representation
 Knowledge
familiarity, awareness, or understanding of someone or
something (facts, rules)

 Awareness or understanding of a circumstance or fact gained

through association or experience.
 Human ability resulting from interpreted information

 Understanding that germinates from combination of data,

information, experience, and individual interpretation.
Knowledge

 Having concepts, procedures, rules, ideas, facts, and associations,

coupled with an ability to use these notions effectively in modeling
different aspects of the world.

 Knowledge may be procedural OR declarative.

 Procedural knowledge is compiled knowledge related to the

programmer e.g., the steps used to solve an algebraic equation are
expressed as procedural knowledge.

 Declarative knowledge, or passive knowledge is expressed as

statement of facts about the world. Personal data in a database is
typical of declarative knowledge. Such data are pieces of
independent knowledge.
 What is Knowledge Representation?

 Knowledge representation (KR) is most fundamentally a surrogate used

to enable an entity to determine consequences by thinking rather than
acting, i.e., by reasoning about the world rather than taking action in it.

 It is a medium for pragmatically efficient computation, i.e., the

computational environment in which thinking is accomplished. One
contribution to this pragmatic efficiency is supplied by the guidance a
representation provides for organizing information so as to facilitate
making the recommended inferences.
 What is Knowledge Representation?

 Knowledge representation is a multidisciplinary subject that applies

theories and techniques from other fields:
1. Logic provides the formal structure and rules of inference.
2. Ontology defines the kinds of things that exist in the application
domain.

 Logic is a formal language for representing facts and properties of a

world in a precise, unambiguous way.
 Propositional Logic
 Predicate Logic

 Logic is concerned with reasoning and the validity of arguments. It is

not concerned with the truth of the sentence, just their validity.
A piece of reason is valid if it leads to a true conclusion
in every situation where the premises are true.

 All men are mortal

 Jack Sparrow is a man
 Therefore, Jack Sparrow is mortal.

This set of statement is considered to be valid because the

conclusion follows logically from the other two statements
(premises)

 All students love exams

 You are a student
 Therefore, you love exams

 Propositional logic is not the study of truth, but of the

relationship between the truth of one statement and that of
another"’ (Hedman 2004)
 Logic is defined by the following:

 Syntax - describes the possible configurations that constitute

sentences.

 Semantics - determines what facts in the world the sentences

refer to i.e. the interpretation. Each sentence makes a claim about
the world.

 Inference Methods - set of rules for generating new sentences

that are necessarily true given that the old sentences are true. The
relationship between sentences is called entailment. The
semantics link these sentences (representation) to facts of the
world. The proof can be used to determine new facts which follow
from the old.
Propositional Logic

 The simplest logic of all. Allows facts about the world to be represented
as sentences formed from:

 Propositional symbols: P, Q, R, S...

 Not: ¬
 And ( ^ )
 Or: ( ˅ )
 Implies:  (if) Connectives
 iff (If and only if): ↔
 Wrapping parentheses: (...)
 logical constants: true, false, unknown
 Quantifiers ∃, ∀

 An expression is referred to as a well-formed-formula (wff) or a

sentence if it is constructed correctly, according to the rules of the
syntax of propositional calculus.
 P ^ Q ˅ (B ^ ¬C)  A ^ B ˅ D ^ (¬E)
Propositional Logic
Atomic propositions
• simple propositions
• represented by a single proposition symbol
• either true or false.

 Earth is round
 It is a hot day
 Ice is cold

A := Earth is round
B := It is a hot day
C := Ice is cold

Compound proposition: Compound propositions are

constructed by combining simpler or atomic propositions, using
parenthesis and logical connectives.
 To use propositional logic, it is first necessary to convert facts and
rules about the real world into logical expressions using the logical
operators
 Let us examine some examples considering simple operators.

 Sentences that use the word ‘and’ in English to express more than
one concept, all of which is true at once, can be easily translated
into logic using the AND operator, ∧. For example:

R:= It is Raining
F:= It is Friday

It is raining and it is Friday might be expressed as:

R∧F
A:= It is hot
B:= The sky is cloudy

A∧B It is hot and the sky is cloudy

AB IF It is hot THEN the sky is cloudy

C:= John can play Tennis

¬ C John cannot play tennis

John can play tennis or football

D:= John can play football
C˅D

If John can play tennis then he can play football

CD
 If you live in France, you can speak French
A:= Live in France
B:= Speak French
AB

 If it is hot and humid, then it is raining

H:= It is hot
M:= It is humid
R:= It is raining
H∧M
H∧MR
Propositional Logic

 It is a useful tool for reasoning, but it has limitation because it

cannot see inside prepositions and take advantage of
relationships among them.

 Paris is the capital of France and Paris has a population of

over two million.
Predicate Logic

 Predicate logic (predicate calculus or First-order logic) is a

powerful language that develops information about the objects
in an easier way and can also express the relationship
between those objects.

 Predicate logic allows us to reason about properties of objects

and the relationships between objects.

 Constants (John, London, Tennis)

 Predicates (father, likes, …)
 Variables (X,Y)
 Quantifier (∃, ∀)
 In predicate calculus, we use predicates to express properties
of objects.

 John is David’s father

 In order to represent a relationship between individual objects,

we can use a predicate specifying the objects as its arguments

 John is David’s father

Father(John, David)

 John likes London

Likes(John, London)

Likes(John, football)
Likes(John, photography)

 Likes(John,?) (what does John like?)

 Likes(John, London)
 Likes(John, football)
 Likes(John, photography)
 Likes(Alice, tennis)
 Likes(Adam, cricket)
 Likes(Sam, tennis)
 Likes(David, photography)

Question: What does John like?

 Likes(John,?)

Question: Who likes photography?

 Likes(?, photography)
- Alison likes Rachel and chocolate
likes(Alison, Rachel)
likes(Alison, Chocolate)

If Rachel and Alison are friends then Alison likes Rachel

Objects: Rachel and Alison

Predicates: friend and like

friend(Alison, Rachel) → likes(Alison, Rachel)

First-order logic

 Universally quantified variables, e.g., ∀ X

 Universal
 for all, for each, for every
 everyone, everything
 All men ….
 ∀X: man(X)

 Existentially quantified variables, e.g., ∃ X

 Existential
 for some, for at least one
 There exists some
 Some men …
 ∃ Y: man(Y)
Using variables with predicates to capture
generalizations

 We can capture generalizations by asserting that any instance

of a given class has the relevant property. For example:

 Every human is mortal

∀x: human(x) → mortal(x)

 Some boys play cricket

 predicate = play(x, y), where x = boy and y = game
∃x: boy(x) ˄ play(x, cricket)

 Some fruits are poisonous

∃ x: fruit(x) ˄ poisonous(x)
Convert to First-order Logic

 Harry, Ron and Draco are students.

 Every student is either wicked or plays Quidditch.
 No one who plays Quidditch likes rain.
 Everyone wicked likes potions.
 Draco dislikes whatever Harry likes and likes whatever Harry
dislikes.
 Draco likes rain and potions.
Predicate Logic - Resolution

 Resolution is a theorem proving technique that works by

building refutation proofs, i.e., proofs by contradictions.
 It was invented by a Mathematician John Alan Robinson in the year
1965.

 Resolution can resolve two clauses if they contain

complementary literals.
 Unification is a key concept in proofs by resolutions
Unification

 Unification: a process of making two different logical atomic

expressions identical by finding a substitution (matching
literals)
 Substitution: when a variable name is replaced by another variable
or element of the domain.
 likes(x,y)
 likes(John, photography) assume/substitute x=John , y=photography

 Two formulas are said to unify if there are possible

assignments of terms to variables that make the formulas in
question identical.
Resolution and CNF

 Resolution is a single rule of inference that can operate

efficiently on a special form of sentences.

 The special form is called Conjunctive Normal Form (CNF) or

clausal form, and has the following properties:

 Every individual sentence is a disjunction (OR) of literals

 Sentence1: likes(a,b) ∨ plays(x,cricket)
 Sentence2: eats(z,pasta) ∨ loves(z,reading)

 These individual sentences are implicitly conjuncted (ANDed)

 Sentence1 ˄ Sentence2
 (likes(a,b) ∨ plays(x,cricket)) ˄ (eats(z,pasta) ∨ loves(z,reading))
Resolution and CNF
Conversion to CNF

 Eliminate all implications →

 Reduce the scope of all ¬ to single term
 Make all variable names unique
 Eliminate Existential Quantifiers
 Eliminate Universal Quantifiers
 Convert to conjunction of disjuncts
 Create separate clause for each conjunct.
Resolution and CNF
Conversion to CNF
Convert to First-order Logic and CNF

 Harry, Ron and Draco are students.

 Every student is either wicked or plays Quidditch.
 No one who plays Quidditch likes rain.
 Everyone wicked likes potions.
 Draco dislikes whatever Harry likes and likes whatever Harry
dislikes.
 Draco likes rain and potions.

 Is Draco wicked?
Apply Resolution Proof Procedure:

 Everyone passing the test is happy.

 Everyone who practices something can pass it.
 John practices everything.
 Everyone who is lucky can pass anything.
 Sam is not lucky but practices everything.

 Prove: John can pass the test.