Knowledge Representation

BCS303 - Artificial Intelligence

 Wumpus world problem
• The Wumpus world problem depicts the value
of a knowledge-based agent and the
interpretation of that knowledge with the help of
reasoning and planning
• The Wumpus world is a 4x4 cave with 16
rooms connected to each other through
passageways. The knowledge-based agent
goes forward in this world.
 Wumpus world problem
• In Wumpus World:
• The cave has a room with a Wumpus, and the
game is over when the agent enters the room;
however, the Wumpus stays in one room.
• The agent is given a single arrow that can be
used to kill the Wumpus.
• There are some “Pits” rooms in the cave, and if
the agent falls in a Pit, they will be stuck there
forever.
 Wumpus world problem
• Goal
• The cave has one room with the possibility of a
heap of gold. The agent’s goal is to find this gold.
• How to “win” it
• The agent has to find the gold and climb out of
the cave without:
• Falling into Pits
• Getting eaten by Wumpus
Additional components to help agent
(sensors)
• Sensors help the agent move forward
successfully.
• Rooms adjacent to the Wumpus have
some stench.
• If the agent is adjacent to a Pit, then he will
perceive a breeze.
• A room with glitter has gold in it.
• Wumpus will scream when it is killed, which can
be heard anywhere in the cave.
• The agent will feel a bump when they hit a wall.
Wumpus world cave problem
 Moves agent can make (actuators)
• The following are the moves an agent can
make once they enter the cave:
• Forward
• Left
• Right
• Grab
• Shoot
 Environment
• The game has:
• A 4x4 grid (16 rooms).
• The agent starts with the room square [1, 1]
and facing towards the right.
• Wumpus and gold can be anywhere except for
the first square [1,1].
• Any room of the cave can be a Pit, except for
the first square.
 Performance measure
• The performance of the agent is measured on
the following basis:
• 1000 reward points if the agent successfully
comes out of the cave with gold
• -1000 points penalty if the agent falls in the Pit
or is eaten by Wumpus
• -1 for each action
• -10 for using an arrow
 Properties of the Wumpus World
• The game is comprised of a single agent.
• The agent can perceive the close environment
that is the room adjacent to them.
• The outcomes of the Wumpus World Problem
are already known.
• The order in the Wumpus World Problem is
important.
• Only the agent can move, so the conditions
stay constant throughout the game.
 Knowledge Representation and
Inference mechanisms
• Inference rules
• Modus ponens
• 𝛼 → 𝛽, 𝛼 ⊢ 𝛽
• Eg. If today is Tuesday, then John will go to
campus (𝛼 → 𝛽), Today is Tuesday (𝛼), Therefore
John will go to campus (𝛽).
• And-Elimination
• 𝛼∧𝛽 ⊢𝛼
• Eg. A star is a sphere of gas and it is held together
by its gravity, therefore a star is a sphere of gas
 More Inference rules
• Commutativity: 𝛼 ∧ 𝛽 ≡ 𝛽 ∧ 𝛼 and 𝛼 ∨ 𝛽 ≡ 𝛽 ∨ 𝛼
• Associativity: (𝛼 ∧ 𝛽) ∧ 𝛾 ≡ 𝛼 ∧ (𝛽 ∧ 𝛾) also on ∨
• Double negation elimination: ¬ ¬𝛼 ≡ 𝛼

• Contrapositive: 𝛼 → 𝛽 ≡ ¬𝛽 → ¬𝛼
• Implication elimination:𝛼 → 𝛽 ≡ ¬𝛼 ∨ 𝛽
• Bicondition elimination:𝛼 ↔ 𝛽 ≡ 𝛼 → 𝛽 ∧ 𝛽 → 𝛼
• De morgan: ¬ 𝛼 ∧ 𝛽 ≡ ¬𝛼 ∨ ¬𝛽 and ¬ 𝛼 ∨ 𝛽 ≡ ¬𝛼 ∧ ¬𝛽

• Distributivity: 𝛼 ∧ 𝛽 ∨ 𝛾 ≡ 𝛼 ∧ 𝛽 ∨ (𝛼 ∨ 𝛽) and 𝛼 ∨
𝛽 ∧ 𝛾 ≡ 𝛼 ∨ 𝛽 ∧ (𝛼 ∨ 𝛽)
 Applying rules to Wumpus
• Consider knowledge base (KB) for Wumpus
• R1: ¬𝑃1,1
• R2: 𝐵1,1 ↔ 𝑃1,2 ∨ 𝑃2,1
• R3: 𝐵2,1 ↔ 𝑃1,1 ∨ 𝑃2,2 ∨ 𝑃3,1
• R4: ¬𝐵1,1
• R5: 𝐵2,1
• Show that ¬𝑃1,2
 Applying rules to Wumpus
• R2: 𝐵1,1 ↔ 𝑃1,2 ∨ 𝑃2,1
• R6: 𝐵1,1 → (𝑃1,2 ∨ 𝑃2,1 ) ∧ (𝑃1,2 ∨ 𝑃2,1 ) → 𝐵1,1
bicondition elimination
• R7: (𝑃1,2 ∨ 𝑃2,1 ) → 𝐵1,1 and elimination
• R8: ¬𝐵1,1 → ¬(𝑃1,2 ∨ 𝑃2,1 ) contrapositive
• Recall R4: ¬𝐵1,1
• R9:¬(𝑃1,2 ∨ 𝑃2,1 ) Modus ponens
• R10: ¬𝑃1,2 ∧ ¬𝑃_2,1 De Morgan
• Gives R11: ¬𝑃1,2 by and elimination
 First-Order Logic

16
 First-order logic
First-order logic (FOL) models the world in terms of
Objects, which are things with individual identities
Properties of objects that distinguish them from others
Relations that hold among sets of objects
Functions, which are a subset of relations where there is
only one “value” for any given “input”
Examples:
Objects: Students, lectures, companies, cars ...
Relations: Brother-of, bigger-than, outside, part-of, has-
color, occurs-after, owns, visits, precedes, ...
Properties: blue, oval, even, large, ...
Functions: father-of, best-friend, second-half, more-than
...
 User provides
Constant symbols representing individuals in the
world
Mary, 3, green
Function symbols, map individuals to individuals
father_of(Mary) = John
color_of(Sky) = Blue
Predicate symbols, map individuals to truth values
greater(5,3)
green(Grass)
color(Grass, Green)
 FOL Provides
Variable symbols
E.g., x, y, foo
Connectives
Same as in propositional logic: not (),
and (), or (), implies (), iff ()
Quantifiers
Universal x or (Ax)
Existential x or (Ex)
 Sentences: built from terms and
atoms
A term (denoting a real-world individual) is a constant
symbol, variable symbol, or n-place function of n
terms.
x and f(x1, ..., xn) are terms, where each xi is a term
A term with no variables is a ground term (i.e., john,
father_of(father_of(john))
An atomic sentence (which has value true or false) is
an n-place predicate of n terms (e.g., green(Kermit))
A complex sentence is formed from atomic sentences
connected by the logical connectives:
P, PQ, PQ, PQ, PQ where P and Q are
sentences
 Sentences: built from terms and
atoms
A quantified sentence adds quantifiers  and 
A well-formed formula (wff) is a sentence
containing no “free” variables. That is, all
variables are “bound” by universal or existential
quantifiers.
(x)P(x,y) has x bound as a universally
quantified variable, but y is free
 Quantifiers
Universal quantification
(x)P(x) means P holds for all values of
x in domain associated with variable
E.g., (x) dolphin(x)  mammal(x)
Existential quantification
( x)P(x) means P holds for some value
of x in domain associated with variable
E.g., ( x) mammal(x)  lays_eggs(x)
Permits one to make a statement about
some object without naming it
 Quantifiers
Universal quantifiers often used with implies to form rules:
(x) student(x)  smart(x) means “All students are smart”
Universal quantification is rarely used to make blanket
statements about every individual in the world:
(x)student(x)  smart(x) means “Everyone in the world is a student
and is smart”
Existential quantifiers are usually used with “and” to specify a
list of properties about an individual:
(x) student(x)  smart(x) means “There is a student who is smart”
Common mistake: represent this EN sentence in FOL as:
(x) student(x)  smart(x)
What does this sentence mean?
 Quantifier Scope
FOL sentences have structure, like programs
In particular, the variables in a sentence have a
scope
For example, suppose we want to say
“everyone who is alive loves someone”
(x) alive(x)  (y) loves(x,y)
Here’s how we scoce the variables
(x) alive(x)  (y) loves(x,y)

Scope of x
Scope of y
 Quantifier Scope
Switching order of universal quantifiers does
not change the meaning
(x)(y)P(x,y) ↔ (y)(x) P(x,y)
“Dogs hate cats”
You can switch order of existential quantifiers
(x)(y)P(x,y) ↔ (y)(x) P(x,y)
“A cat killed a dog”
Switching order of universals and existentials
does change meaning:
Everyone likes someone: (x)(y) likes(x,y)
Someone is liked by everyone: (y)(x) likes(x,y)
 Connections between All and Exists
We can relate sentences involving  and  using De Morgan’s
laws:
1. (x) P(x) ↔ (x) P(x)
2. (x) P ↔ (x) P(x)
3. (x) P(x) ↔  (x) P(x)
4. (x) P(x) ↔ (x) P(x)
Examples
1. All dogs don’t like cats ↔ No dogs like cats
2. Not all dogs dance ↔ There is a dog that doesn’t dance
3. All dogs sleep ↔ There is no dog that doesn’t sleep
4. There is a dog that talks ↔ Not all dogs can’t talk
 Quantified inference rules
Universal instantiation
x P(x)  P(A)
Universal generalization
P(A)  P(B) …  x P(x)
Existential instantiation
x P(x) P(F) skolem constant F
F must be a “new” constant not
appearing in the KB
Existential generalization
P(A)  x P(x)
 Universal instantiation
(a.k.a. universal elimination)

If (x) P(x) is true, then P(C) is true, where

C is any constant in the domain of x, e.g.:
(x) eats(John, x) 
eats(John, Cheese18)
Note that function applied to ground terms
is also a constant
(x) eats(John, x) 
eats(John, contents(Box42))
 Existential instantiation
(a.k.a. existential elimination)
From (x) P(x) infer P(c), e.g.:
(x) eats(Mickey, x)  eats(Mickey, Stuff345)
The variable is replaced by a brand-new constant not
occurring in this or any sentence in the KB
Also known as skolemization; constant is a skolem
constant
We don’t want to accidentally draw other inferences
about it by introducing the constant
Can use this to reason about unknown objects, rather
than constantly manipulating existential quantifiers
 Existential generalization
(a.k.a. existential introduction)
If P(c) is true, then (x) P(x) is inferred, e.g.:
Eats(Mickey, Cheese18) 
(x) eats(Mickey, x)
All instances of the given constant symbol
are replaced by the new variable symbol
Note that the variable symbol cannot already
exist anywhere in the expression
 Translating English to FOL
Every gardener likes the sun
x gardener(x)  likes(x,Sun)
You can fool some of the people all of the time
x t person(x)  time(t)  can-fool(x, t)
You can fool all of the people some of the time
x t (person(x)  time(t) can-fool(x, t)) Note 2 possible
x (person(x)  t (time(t) can-fool(x, t)) readings of NL
sentence
All purple mushrooms are poisonous
x (mushroom(x)  purple(x))  poisonous(x)
 Translating English to FOL

No purple mushroom is poisonous (two ways)

x purple(x)  mushroom(x)  poisonous(x)
x (mushroom(x)  purple(x))  poisonous(x)

There are exactly two purple mushrooms

x y mushroom(x)  purple(x)  mushroom(y) 
purple(y) ^ (x=y)  z (mushroom(z) 
purple(z))  ((x=z)  (y=z))

Obana is not short

short(Obama)
 Example: A simple genealogy KB by FOL
Build a small genealogy knowledge base using FOL that
contains facts of immediate family relations (spouses, parents, etc.)
contains definitions of more complex relations (ancestors, relatives)
is able to answer queries about relationships between people
Predicates:
parent(x, y), child(x, y), father(x, y), daughter(x, y), etc.
spouse(x, y), husband(x, y), wife(x,y)
ancestor(x, y), descendant(x, y)
male(x), female(y)
relative(x, y)
Facts:
husband(Joe, Mary), son(Fred, Joe)
spouse(John, Nancy), male(John), son(Mark, Nancy)
father(Jack, Nancy), daughter(Linda, Jack)
daughter(Liz, Linda)
etc.
Rules for genealogical relations
(x,y) parent(x, y) ↔ child (y, x)
(x,y) father(x, y) ↔ parent(x, y)  male(x) ;similarly for
mother(x, y)
(x,y) daughter(x, y) ↔ child(x, y)  female(x) ;similarly for
son(x, y)
(x,y) husband(x, y) ↔ spouse(x, y)  male(x) ;similarly for
wife(x, y)
(x,y) spouse(x, y) ↔ spouse(y, x) ;spouse relation is
symmetric
(x,y) parent(x, y)  ancestor(x, y)
(x,y)(z) parent(x, z)  ancestor(z, y)  ancestor(x, y)
(x,y) descendant(x, y) ↔ ancestor(y, x)
(x,y)(z) ancestor(z, x)  ancestor(z, y)  relative(x, y)
;related by common ancestry
(x,y) spouse(x, y)  relative(x, y) ;related by marriage
(x,y)(z) relative(z, x)  relative(z, y)  relative(x, y)
;transitive
(x,y) relative(x, y) ↔ relative(y, x) ;symmetric
Queries
ancestor(Jack, Fred) ; the answer is
yes
relative(Liz, Joe) ; the answer is
yes
relative(Nancy, Matthew)
;no answer in general, no if
under closed world assumption
(z) ancestor(z, Fred)  ancestor(z,
Liz)
 Axioms for Set Theory in FOL
1. The only sets are the empty set and those made by
adjoining something to a set:
s set(s) <=> (s=EmptySet) v (x,r Set(r) ^ s=Adjoin(s,r))
2. The empty set has no elements adjoined to it:
~ x,s Adjoin(x,s)=EmptySet
3. Adjoining an element already in the set has no effect:
x,s Member(x,s) <=> s=Adjoin(x,s)
4. The only members of a set are the elements that were
adjoined into it:
x,s Member(x,s) <=> y,r (s=Adjoin(y,r) ^ (x=y 
Member(x,r)))
 Axioms for Set Theory in FOL
5. A set is a subset of another iff all of the 1st set’s
members are members of the 2nd:
s,r Subset(s,r) <=> (x Member(x,s) => Member(x,r))
6. Two sets are equal iff each is a subset of the other:
s,r (s=r) <=> (subset(s,r) ^ subset(r,s))
7. Intersection
x,s1,s2 member(X,intersection(S1,S2)) <=> member(X,s1) ^
member(X,s2)
8. Union
x,s1,s2 member(X,union(s1,s2)) <=> member(X,s1) 
member(X,s2)
 Semantics of FOL
Domain M: the set of all objects in the world (of interest)
Interpretation I: includes
Assign each constant to an object in M
Define each function of n arguments as a mapping Mn => M
Define each predicate of n arguments as a mapping Mn => {T, F}
Therefore, every ground predicate with any instantiation will have
a truth value
In general there is an infinite number of interpretations because
|M| is infinite
Define logical connectives: ~, ^, v, =>, <=> as in PL
Define semantics of (x) and (x)
(x) P(x) is true iff P(x) is true under all interpretations
(x) P(x) is true iff P(x) is true under some interpretation
Model: an interpretation of a set of sentences
such that every sentence is True
A sentence is
satisfiable if it is true under some interpretation
valid if it is true under all possible
interpretations
inconsistent if there does not exist any
interpretation under which the sentence is true
Logical consequence: S |= X if all models of S
are also models of X
 Axioms, definitions and theorems
Axioms are facts and rules that attempt to capture all of
the (important) facts and concepts about a domain;
axioms can be used to prove theorems
Mathematicians don’t want any unnecessary (dependent) axioms,
i.e. ones that can be derived from other axioms
Dependent axioms can make reasoning faster, however
Choosing a good set of axioms for a domain is a design problem
A definition of a predicate is of the form “p(X) ↔ …” and
can be decomposed into two parts
Necessary description: “p(x)  …”
Sufficient description “p(x)  …”
Some concepts don’t have complete definitions (e.g., person(x))
 More on definitions
Example: define father(x, y) by parent(x, y) and
male(x)
parent(x, y) is a necessary (but not sufficient)
description of father(x, y)
father(x, y)  parent(x, y)
parent(x, y) ^ male(x) ^ age(x, 35) is a sufficient (but
not necessary) description of father(x, y):
father(x, y)  parent(x, y) ^ male(x) ^ age(x, 35)
parent(x, y) ^ male(x) is a necessary and sufficient
description of father(x, y)
parent(x, y) ^ male(x) ↔ father(x, y)
 More on definitions

S(x) is a P(x)
necessary (x) P(x) => S(x)
S(x)
condition of P(x)

S(x) is a S(x)
sufficient (x) P(x) <= S(x)
P(x)
condition of P(x)

S(x) is a P(x)
necessary and (x) P(x) <=> S(x)
sufficient S(x)
condition of P(x)
 Higher-order logic
FOL only allows to quantify over variables, and
variables can only range over objects.
HOL allows us to quantify over relations
Example: (quantify over functions)
“two functions are equal iff they produce the same
value for all arguments”
f g (f = g)  (x f(x) = g(x))
Example: (quantify over predicates)
r transitive( r )  (xyz) r(x,y)  r(y,z)  r(x,z))
More expressive, but undecidable, in general
 Expressing uniqueness
Sometimes we want to say that there is a single,
unique object that satisfies a certain condition
“There exists a unique x such that king(x) is true”
x king(x)  y (king(y)  x=y)
x king(x)  y (king(y)  xy)
! x king(x)
“Every country has exactly one ruler”
c country(c)  ! r ruler(c,r)
Iota operator: “ x P(x)” means “the unique x such that
p(x) is true”
54
“The unique ruler of Freedonia is dead”
dead( x ruler(freedonia,x))
 Notational differences
Different symbols for and, or, not, implies, ...
        
p v (q ^ r)
p + (q * r) etc
Prolog
cat(X) :- furry(X), meows (X), has(X, claws)
Lispy notations
(forall ?x (implies (and (furry ?x)
(meows ?x)
(has ?x claws))
55
(cat ?x)))
 Summary
First order logic (FOL) introduces predicates,
functions and quantifiers
Much more expressive, but reasoning is more
complex
Reasoning is semi-decidable
FOL is a common AI knowledge representation
language
FOL variables range over objects
HOL variables can ranger over functions, predicates
or sentences