Artificial Intelligence

(2180703)

Unit-3 Knowledge Representation Issues

Rutal Mahajan
S.N.P.I.T.R.C., Umrakh
[email protected]

Knowledge representation Issues (AI B.E.8th semester)

2/23/2019 1
by Rutal Mahajan
Why knowledge Representation

To solve complex problems we need:

1. Large amount of knowledge
2. Mechanism for representation and manipulation of
existing knowledge to create new solution.

Knowledge Representation
Facts: Things we want to represent. Truth in some
relevant world.
Representation of facts.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 2
by Rutal Mahajan
Knowledge and Knowledge Model
• Knowledge is a general term.
• Knowledge is a progression that starts with data which is of limited
utility.
• By organizing or analyzing the data, we understand what the
data means, and this becomes information.
• The interpretation or evaluation of information yield knowledge.
• An understanding of the principles embodied within the
knowledge is wisdom.

Fig. Knowledge progression

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 3
by Rutal Mahajan
Knowledge and Knowledge Model

• Data is viewed as collection of disconnected facts.

Example : It is raining.

• Information emerges when relationships among facts are established and understood;
Provides answers to "who", "what", "where", and "when".
Example : The temperature dropped 15 degrees and then it started raining.

• Knowledge emerges when relationships among patterns are identified and understood;
Provides answers as "how".
Example : If the humidity is very high and the temperature drops substantially, then
atmospheres is unlikely to hold the moisture, so it rains.

• Wisdom is the pinnacle of understanding, uncovers the principles of relationships that

describe patterns. Provides answers as "why" .
Example : Encompasses understanding of all the interactions that happen between raining,
evaporation, air currents, temperature gradients, changes, and raining.
Knowledge Representation Issues (AI B.E.8th semester)
2/23/2019 4
by Rutal Mahajan
Knowledge and Knowledge Model

• The model represents transitions and understanding.

• the transitions are from data, to information, to knowledge, and finally to wisdom;
• the understanding support the transitions from one stage to the next stage.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 5
by Rutal Mahajan
Knowledge Types
• Knowledge is categorized into two major types: Tacit and Explicit
• term “Tacit” corresponds to informal or implicit type of knowledge,
• term “Explicit” corresponds to formal type of knowledge.

Tacit knowledge Explicit knowledge

Exists within a human being; Exists outside a human being;
it is embodied. it is embedded
Difficult to articulate formally Can be articulated formally.
Difficult to share/communicate Can be shared, copied, processed and
stored.
Hard to steal or copy. Easy to steal or copy
Drawn from experience, action, Drawn from artifact of some type as
subjective insight. principle, procedure, process, concepts.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 6
by Rutal Mahajan
Knowledge Types
• Cognitive psychologists sort knowledge into Declarative and Procedural

Procedural knowledge Declarative knowledge

examples : procedures, rules, example : concepts, objects, facts,
strategies, agendas, models. propositions, assertions, semantic
nets, logic and descriptive models.
focuses on tasks that must be refers to representations of objects
performed to reach a particular and events; knowledge about facts
objective or goal. and relationships
Knowledge about "how to do Knowledge about "that something
something"; e.g., to determine if is true or false". e.g., A car has
Peter or Robert is older, first find four tyres; Peter is older than
their ages. Robert;

All declarative knowledge are explicit knowledge;

it is knowledge that can be and has been articulated.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 7
by Rutal Mahajan
Relationships among Knowledge Types
• declarative knowledge is tied to
"describing"
• procedural knowledge is tied to
"doing."

Therefore, we may view,

− all procedural knowledge as tacit,
− all declarative knowledge as explicit.

The arrows connecting explicit with declarative and tacit with procedural,
indicate the strong relationships exist among them.
− The arrow connecting declarative and procedural indicates that we often
develop procedural knowledge as a result of starting with declarative
knowledge. i.e., we often "know about" before we "know how".

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 8
by Rutal Mahajan
Why Knowledge Representation ??

The Problem solving requires

− formal knowledge representation, and
− conversion of informal (implicit) knowledge to formal (explicit) knowledge.

The Knowledge and the Representation are distinct entities, play a central but
distinguishable roles in intelligent system.
− Knowledge is a description of the world;
it determines a system's competence by what it knows.
− Representation is the way knowledge is encoded;
it defines the system's performance in doing something.

A good representation enables fast and accurate access to knowledge

and understanding of the content.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 9
by Rutal Mahajan
Knowledge Representation Framework
Computer requires well defined problem description.
first to formulate description in our spoken language and
then represent it in formal language so that computer can understand. The computer can then
use an algorithm to compute an answer. This process is illustrated below.

Implicit
knowledge

Explicit
knowledge

The steps are

• The informal formalism of the problem takes place first.
• It is then represented formally and the computer produces an output.
• This output can then be represented in a informally described solution that user
understands or checks for consistency.
Knowledge Representation Issues (AI B.E.8th semester)
2/23/2019 10
by Rutal Mahajan
Facts Representation and Mapping
Knowledge is a collection of “facts” from some domain.
We need a representation of facts that can be manipulated by a program.
Normal English is insufficient, too hard currently for a computer program to draw inferences in
natural languages.
Thus some symbolic representation is necessary.

Therefore, we must be able to map "facts

to symbols" and "symbols to facts" using
forward and backward representation
mapping.

Example : Consider an English sentence

Facts Representations
• Spot is a dog A fact represented in English sentence
• dog (Spot) Using forward mapping function the above fact is represented in logic
• ∀ x : dog(x) → hastail (x) A logical representation of the fact that "all dogs have tails"
Now using deductive mechanism we can generate a new representation of object:
• hastail (Spot) A new object representation
• Spot has a tail Using backward mapping function to generate English sentence
[it is new knowledge]
Knowledge Representation Issues (AI B.E.8th semester)
2/23/2019 11
by Rutal Mahajan
Forward and Backward Representation
The forward and backward representations are elaborated below :

The Knowledge and the Representation are

distinct entities, play a central but
distinguishable roles in intelligent system.

− Knowledge is a description of the world;

it determines a system's competence by what
it knows.
− Representation is the way knowledge is
encoded;
it defines the system's performance in doing
something.

• The doted line on top indicates the abstract reasoning process that a program is intended to
model.
• The solid lines on bottom indicates the concrete reasoning process that the program
performs.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 12
by Rutal Mahajan
Knowledge Representation System Requirements / Properties

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 13
by Rutal Mahajan
Approaches to Knowledge Representation

Knowledge can be represented using the following approaches:

1. Logical Representation
1. Propositional Logic
2. First Order Predicate Logic
3. Fuzzy Logic
4. Other Logic

2. Production rule Representations

3. Semantic Networks

4. Frame Representation

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 14
by Rutal Mahajan
Issues in Knowledge Representation
(refer pgno.87-95 of Elain rich & Kevin Knight-3rd edition for detailed understanding)
1. Important Attributes: Any attribute of objects so basic that they occur in almost every
problem domain ?
(There can be attributes that occur in many different types of problem with different names)
For example, instance and isa , are important because each supports property inheritance.

2. Relationships: Any important relationship that exists among object attributes ?

The relationships such as inverse, existence; among various attributes of an object need to
be represented without ambiguity.

3. Granularity: At what level of detail should the knowledge be represented ?

E.g Tom feed dog vs Tom gives dog a bone
we need to add inferential rule that give(x, food)-> feed(x)

4. Set of Objects: how set of objects be represented?

5. Finding Right Structure: Given a large amount of knowledge stored, how can relevant
parts be accessed?

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 15
by Rutal Mahajan
Inheritable Knowledge Representation

• The directed arrows represent attributes (isa, instance, and team) originating at the object
being described and terminating at the object or its value.
• the box nodes represents objects and values of the attributes.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 16
by Rutal Mahajan
Issues in Knowledge Representation

• The directed arrows

represent attributes
(isa, instance, and team).
• the box nodes represents
objects and values of the
attributes.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 17
by Rutal Mahajan
Logic Representation
1. What is Logic?
2. What is PL?
3. Examples of PL
4. Inference Rules
5. Advantage and Disadvantage of PL

Logic:
• way of representing facts formally
• formally representing facts is important (logical representation) because…
it is powerful way of deriving new knowledge from old.
• E.g: New statement is proved true by proving that it follows the
statements already known - Mathematical induction

• Logic includes: syntax, semantics and inference procedure

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 18
by Rutal Mahajan
Logic Representation

Logic:
• Syntax: specifies symbols in language about how they can be combined to
form sentences
• Semantics: it specifies how to assign a truth value to a sentence based on
its meaning in the world.
• Inference Procedure: specifies methods for computing new sentences from
an existing sentences.
E.g. Induction or Deduction

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 19
by Rutal Mahajan
Propositional Logic Representation
Types of Logic: Propositional logic, predicate logic, temporal logic,
description logic etc.

• Propositional logic- is the study of statements and their connectivity

• Predicate logic- is the study of individuals and their properties.

Propositional Logic:
Proposition is a statement, in which English would be a declarative sentence.

Examples:
(a) sky is blue
(b) Snow is cold
(c) 12*12=144

Propositions are “sentences”, either TRUE or FALSE but not both.

It is fundamental to all logics

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 20
by Rutal Mahajan
Propositional Logic Examples

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 21
by Rutal Mahajan
Propositional Logic -Terminologies
Statement: sentence that can be true/false.

Properties of statement:
• Tautology/Satisfiability: means valid sentence. a sentence
which is true for all the interpretations.
Eg.”we wear woollen cloths”
“It is Hot or It is not Hot” (A V ~A) (A or Not A)

• Contradiction: if there is no interpretation for which sentence is

true.
Eg. “Japan is capital of India”
“It is Hot And It is not Hot” (A ^ ~A) (A and Not A)

• Entailment/ Validity : a sentence is valid if it is true for every

interpretation. X|=Y (i.e. X entails Y , means whenever sentence X
is true Sentence Y will be true)
Eg. X= Priya is Pooja’s Mother’s sister
Y=Priya is Pooja’s Aunt.
Then X|=Y ( X entails Y)

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 22
by Rutal Mahajan
Propositional Logic –Inference Rules

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 23
by Rutal Mahajan
Propositional Logic –Inference Rules
1. Idempotent rule: 7. Bidirectional Implication elimination:
P ˄ P ==> P ( P  Q ) ==> ( P  Q ) ˄ (Q  P)
P ˅ P ==> P
8. Contrapositive rule:
2. Commutative rule: P  Q => ‫ך‬P  ‫ך‬Q
P ˄ Q ==> Q ˄ P 9. Double Negation rule:
P ˅ Q ==> Q ˅ P
‫ך (ך‬P) => P
3.Associative rule: 10. Absorption Rule:
P ˄ (Q ˄ R) ==> (P ˄ Q) ˄ R P ˅ ( P ˄ Q) => P
P ˅ (Q ˅ R) ==> (P ˅ Q) ˅ R P ˄ ( P ˅ Q) => P
4. Distributive Rule: 11.Fundamental identities:
P ˅ (Q ˄ R) ==> (P ˅ Q) ˄ (P ˅ R) P ˄ ‫ ך‬p => F [contradiction]
P ˄ (Q ˅ R) ==> (P ˄ Q) ˅ (P ˄ R) P ˅ ‫ך‬P => T [Tautology]

5. De-Morgan’s Rule:
‫(ך‬P ˅ Q) ==> ‫ך‬P ˄ ‫ך‬Q P ˅ T => P
‫( ך‬P ˄ Q) ==> ‫ ך‬P ˅ ‫ך‬Q P ˅ F => P
P ˅ ‫ ך‬T => P
6. Implication elimination:
P  Q => ‫ך‬P ˅ Q
P ˄ F => F
P ˄ T => P
Knowledge Representation Issues (AI B.E.8th semester)
2/23/2019 24
by Rutal Mahajan
Propositional Logic –Representation and use
Example 1:
“I will get wet if it rains and I go out of the house”

Let Propositions be:

W : “I will get wet “
R : “it rains “
S : “I go out of the house”
(S ˄ R)  W

Example 2:
“If it is raining, then it is not sunny”

RAINING : It is raining

SUNNY :It is sunny

WINDY: It is windy
If it is raining, then it is not sunny
RAINING  SUNNY

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 25
by Rutal Mahajan
Propositional Logic –Normal Forms
1. Conjunctive normal form (CNF):
e.g. ( P ˅ Q ˅ R ) ˄ (P ˅ Q ) ˄ (P ˅ R ) ˄ P
It is conjunction (˄) of disjunctions (˅)

Where disjunctions are:

1. ( P ˅ Q ˅ R )
2. (P ˅ Q )
3. (P ˅ R ) clauses
4. P

2. Disjunctive normal form (DNF):

e.g. ( P ˄ Q ˄ R ) ˅ (P ˄ Q ) ˅ (P ˄ R ) ˅ P

It is disjunction (˅) of conjunctions (˄)

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 26
by Rutal Mahajan
Propositional Logic –Procedure to convert to Normal Forms

1. Eliminate implications and biconditionals using formulas:

( P  Q ) ==> ( P  Q ) ˄ (Q  P)
P  Q => ‫ך‬P ˅ Q

2. Apply De-Morgan’s Law and reduce NOT symbols so as to bring negations

before the atoms. Use:
‫(ך‬P ˅ Q) ==> ‫ך‬P ˄ ‫ך‬Q
‫( ך‬P ˄ Q) ==> ‫ ך‬P ˅ ‫ך‬Q

3. Use distributive and other laws & equivalent formulas to obtain Normal forms.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 27
by Rutal Mahajan
Propositional Logic –CNF example

Q. Convert into CNF : ( ( PQ )R )

Solution:

Step 1: ( ( PQ )R ) ==> ( ( ‫ך‬P ˅ Q)R)

==> ‫ך ( ך‬P ˅ Q) ˅ R

Step 2: ‫ך ( ך‬P ˅ Q) ˅ R ==> (P ˄ ‫ ך‬Q ) ˅ R

Step 3: (P ˄ ‫ ך‬Q ) ˅ R ==> ( P ˅ R ) ˄ (‫ ך‬Q ˅ R )

CNF

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 28
by Rutal Mahajan
Propositional Logic –How it works?? Proof by Resolution
E.g. “ If it is Hot then it is Humid. If it is humid then it will rain. It is hot.”
prove that “ it will rain.”

Solution:
Let us denote these statements with propositions H,O and R:
H: “ It is humid”.
O: “ It is Hot”.
R: “It will rain”.

Formulas corresponding to the sentences are:

1. “if it is hot then it is humid” [ OH] ==> ‫ך‬O ˅ H
2. “If it is humid then it will rain”. [ HR] ==> ‫ך‬H ˅ R
3. “ It is Hot” [O] ==> O

To prove: R.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 29
by Rutal Mahajan
Propositional Logic –How it works?? Proof by Resolution
Let us assume “it will NOT rain” [ ‫ך‬R ]
“if it is hot then it is humid” [ OH] ==> ‫ך‬O ˅ H
“If it is humid then it will rain”. [ HR] ==> ‫ך‬H ˅ R
[ ‫ך‬R ] [ ‫ך‬H ˅ R] “ It is Hot” [O] ==> O

‫ך‬H [ ‫ך‬O ˅ H]

‫ך‬O O

E
[EMPTY CLAUSE / CONTRADICTION ]
Knowledge Representation Issues (AI B.E.8th semester)
2/23/2019 30
by Rutal Mahajan
Propositional Logic –How it works?? Proof by Resolution
Since an empty clause ( E ) has been deduced we say that our assumption is wrong
and hence we have proved: “It will rain”

Using Prepositional Logic:

• Theorem proving is decidable BUT (we got yes/ no answer-It will rain)

• It Cannot represent objects and quantification.

• Hence we go for PREDICATE LOGIC –Unit 4

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 31
by Rutal Mahajan
Propositional Logic Advantage and Disadvantage
Advantage:
1.Simple knowledge representation language
2. Sufficient and efficient technique for solving some AI problems
3. Foundation for higher logics like FOL-First order Logic

Disadvantage:
1. Can not express complex AI problems
2. Weak knowledge representation language.
- hard to identify individual entities
E.g. Pooja, Priya, Mumbai etc.
- cannot represent properties of individual entities
E.g. Pooja is Tall
- cannot express specialization, generalization or patterns
E.g. All rectangles have 4 sides
Knowledge Representation Issues (AI B.E.8th semester)
2/23/2019 32
by Rutal Mahajan
Questions from GTU( last updated winter 2018)
1) Explain the different issues in Knowledge representation [7 Marks](Nov-18,Dec-14,May-16]
2) Explain mapping between facts and representations in knowledge representation.
3) Explain property inheritance algorithm with example.
4) Discuss approaches to knowledge representation
5) Differentiate tactic and explicit knowledge
6) What are various resolution strategies? Explain any.
7) What is resolution? Explain algorithm for resolution in propositional logic
8) Differentiate procedural and declarative knowledge representation mapping.
9) Explain Horn Clause in detail.

Knowledge Representation Issues (AI B.E.8th semester)

2/23/2019 33
by Rutal Mahajan
 References

1. Artificial Intelligence, 3rd Edition, Elaine Rich, Kevin Knight,

Shivshankar B Nair

What is AI? (AI B.E.8th semester)

2/23/2019 34
by Rutal Mahajan