Artificial Intelligence

CS 165A

 First order logic (Ch. 8)

1
Announcement
• Homework 4 will be announced today. Due Dec 3
– It will cover the lecture on Nov 26.
• Midterm Statistics

2
Propositional logic vs. FOL
• Propositional logic:
– P stands for “All men are mortal”
– Q stands for “Tom is a man”
– What can you infer from P and Q?
 Nothing!

• First-order logic:
– x Man (x)  Mortal (x)
– Man (Tom)
– What can you infer from these?
 Can infer Mortal (Tom)

5
First-Order Logic (FOL)
• An extension to propositional logic in which quantifiers
can bind variables in sentences
– Universal quantifier (  )
– Existential quantifier (  )
– Variables: x, y, z, a, joe, table…

• Examples
–  x Beautiful (x)
–  x Beautiful (x)

6
First-Order Logic (cont.)
• It is by far the most studied and best understood logic in
use
• It does have limits, however
– Quantifiers ( and ) can only be applied to objects, not to
functions or predicates
 Cannot write P P (mom) = good

 This is why it’s called first-order

– This limits its expressiveness

• Let’s look at the syntax of first-order logic
– I.e., what logical expressions can you legally construct?

7
 FOL Syntax
• Symbols
– Object symbols (constants): P, Q, Fred, Desk, True, False, …
 These refer to things
– Predicate symbols: Heavy, Smart, Mother, …
 These are true or false statements about objects: Smart(rock)
– Function symbols: Cosine, IQ, MotherOf, …
 These return objects, exposing relations: IQ(rock)
– Variables: x, y, , …
 These represent unspecified objects
– Logical connectives to construct complex sentences: , , , , 
– Quantifiers:  (universal),  (existential)
– Equality: =
• Usually variables will be lower-case, other symbols capitalized
8
FOL Syntax (cont.)
• Terms
– Logical expressions that refer to objects (evaluates to an object)
– Can be constants, variables, functions
• Examples
– P
– 2001
– Richard
– x
– y
– BrotherOf (Richard)
– Age (NephewOf (x)) [Why not AgeOf ( ) ? (No reason...!)]

Remember – syntax and semantics are different, and separate!!

9
FOL Syntax
• Note on predicates and functions: typical usage
– Beautiful (y)  “y is beautiful”
– Mother (x)  “x is a Mother”
– BrotherOf (x, y)  “x is a brother of y” Predicates
– NextTo (x, y)  “x is next to y”

– BrotherOf (x)  “the brother of x”

– NextTo (y)  “the thing next to y” Functions
– SquareRoot (x)  “the square root of x”

10
FOL Sentences
Constant, variable, or
1. Atomic sentence function – evaluates to an
– A predicate applied to some terms object
 Brothers (Bill, FatherOf (John))
 LessThan (3, 5)
– Equality – states that two terms refer to the same object
 x = MotherOf (y)
 Instructor (cs165a) = Turk
 This is equivalent to the predicate: Equal (Instructor (cs165a), Turk)

2. Logical (complex) sentence – logical combination of other sentences

– Brothers (Bill, HusbandOf (Sue))
– Above (Sky, Ground)  Below (Ground, Sky)
– Brothers (Bill, John)  Brothers (John, Bill)
3. Quantified sentence – sentences with quantified variables
–  x,y ParentOf (x, y)  ChildOf (y, x)
–  x US-President (x)

11
Universal Quanitifer (“For all…”)
•  <variables> <sentence>
–  x – “For all x…”
–  x, y – “For all x and y…”
• Examples
– “Everything is beautiful”
  x Beautiful (x)
 Equivalent to: i Beautiful (xi)
– Beautiful (Joe)  Beautiful (Mary)  Beautiful (apple) 
Beautiful (dirt)  Beautiful (death)  …
– “All men are mortal”
  x Man (x)  Mortal (x)
– “Everyone in the class is smart”
  x Enrolled (x, cs165a)  Smart (x)
– What does this mean:
  x Enrolled (x, cs165a)  Smart (x)

12
Expansion of universal quantifier
•  x Enrolled(x, cs165a)  Smart(x)
• This is equivalent to
– Enrolled (Tom, cs165a)  Smart (Tom) 
Enrolled (Mary, cs165a)  Smart (Mary) 
Enrolled (Chris, cs165a)  Smart (Chris) 
Enrolled (chair, cs165a)  Smart (chair) 
Enrolled (dirt, cs165a)  Smart (dirt) 
Enrolled (surfboard, cs165a)  Smart (surfboard) 
Enrolled (tooth, cs165a)  Smart (tooth) 
Enrolled (Mars, cs165a)  Smart (Mars)  …
– Everything!
• So,  x Enrolled (x, cs165a)  Smart (x) is equivalent to
– Enrolled (Tom, cs165a)  Smart (Tom) 
Enrolled (chair, cs165a)  Smart (chair)  …
13
Existential Quantifier (“There exists…”)
•  <variables> <sentence>
–  x – “There exists an x such that…”
–  x, y – “There exist x and y such that…”
• Examples
– “Somebody likes me”
  x Likes (x, Me) ???

 Equivalent to: i Likes (xi, Me)

– Likes (Joe, Me)  Likes (Mary, Me)  Likes (apple, Me) 

Likes (dirt, Me)  Likes (death, Me)  …
 Really “Something likes me”

–  x Person (x)  Likes (x, Me)

–  x Enrolled (x, cs165a)  WillReceiveAnA+ (x)

14
Scope of Quantifiers
• Scope – the portion of the {program, function, definition,
sentence…} in which the object can be referred to by its
simple name
• Parentheses can clarify the scope (make it explicit)
–  x (  y <sentence> )
• However, the scope of quantifiers is often implicit
–  w  x  y  z <sentence>
is the same as
–  w ( x ( y ( z <sentence>) ) )

Usually the scope of a quantifier extends as far right as possible

–  w  x  y  z <term-1>  <term-2>
is the same as
–  w  x  y  z (<term-1>  <term-2>)
15
Scope of Quantifiers (cont.)
Equivalent
•  x <sentence-1>   x <sentence-2>
–  x (<sentence-1>)   x (<sentence-2>)
–  x (<sentence-1>)   y (<sentence-2-subst-y-for-x>)
–  x Rich(x)  Beautiful(x)
 “Someone is both rich and beautiful”

–  x Rich(x)   x Beautiful(x)
 “Someone is rich and someone is beautiful”

 Same as  x Rich(x)   y Beautiful(y)

• How about
–  x ( Rich(x)   x (Beautiful(x)) )
– The same as  x Rich(x)   x Beautiful(x)

Same as in scope of variables in programming (C/C++, Java, etc.)

16
Order, nesting of Quantifiers
• Implied nesting:
–  x  y <sentence> is the same as  x ( y <sentence>)
–  x  y <sentence> is the same as  x ( y <sentence>)

•  x  y <sentence> is the same as  y  x <sentence>

– Also,  x, y <sentence>

•  x  y <sentence> is the same as  y  x <sentence>

– Also,  x, y <sentence>

•  x  y <sentence> is not the same as  y  x <sentence>

– Try  x  y Loves(x, y) and  y  x Loves(x, y)

17
Example of quantifier order
•  x  y Loves(x, y)
–  x [  y Loves(x, y) ]
–  x [ Loves(x, Fred)  Loves(x, Mary)  Loves(x, Chris)  … ]
– “There is at least one person who loves everybody”
 Assuming the domain consists of only people

•  y  x Loves(x, y)
–  y [  x Loves(x, y) ]
–  y [ Loves(Joe, y)  Loves(Sue, y)  Loves(Kim, y)  … ]
– “Everybody is loved by at least one person”

18
Logical equivalences about  and 
•  can be expressed using 
–  x Statement-about-x … is equivalent to …
–  x Statement-about-x
– Example:  x Likes(x, IceCream)
  x Likes(x, IceCream)

•  can be expressed using 

–  x Statement-about-x … is equivalent to …
–  x Statement-about-x
– Example:  x Likes(x, Spinach)
  x  Likes(x, Spinach)

19
Examples of FOL
• Brothers are siblings
–  x, y Brother(x, y)  Sibling(x, y)
• Sibling is transitive
–  x, y, z Sibling(x, y)  Sibling(y, z)  Sibling(x, z)
• One’s mother is one’s sibling’s mother
–  x, y, z Mother(x, y)  Sibling(y, z)  Mother(x, z)
• A first cousin is a child of a parent’s sibling
–  x, y FirstCousin(x, y) 
 v, w Parent(v, x)  Sibling(v, w)  Parent(w, y)

20
Implication and Equivalence
• Note the difference between  and 
– Implication / conditional (  )
 A  B : “A implies B”, “If A then B”
– Equivalence / biconditional (  )
 A  B : “A is equivalent to B”
 Same as (A  B)  (B  A) : “A if and only if B”, “A iff B”

• For “Sisters are siblings”, which one?

 x, y Sister(x, y)  Sibling(x, y)
 x, y Sister(x, y)  Sibling(x, y)

21
Where we are…
• Basics of logic: Propositional logic

• More general logic representation: First-order logic

• Now, let’s see how to use FOL to do logical inference

– I.e., to reason about the world

22
George Boole (1815-1864)
• More than 100 years later, he didn’t know
about Leibniz, but proceeded to bring to life
part of Leibniz’ dream
• His insight: Logical relationships are
expressible as a kind of algebra
– Letters represent classes (rather than numbers)
– So logic can be viewed as a form of
mathematics
• Published The Laws of Thought

• He extended Aristotle's simple syllogisms to a broader

range of reasoning
– Syllogism: Premise_1, Premise_2  Conclusion
– His logic: Propositional logic
Geoffrey Hinton is the great-great-grandson of the mathematician
and educator Mary Everest Boole and her husband, the logician George Boole.
23
Gottlob Frege (1848-1925)
• He provided the first fully developed system of
logic that encompassed all of the deductive
reasoning in ordinary mathematics.
• He intended for logic to be the foundation of
mathematics – all of mathematics could be
based on, and derived from, logic
• In 1879 he published Begriffsschrift, subtitled
“A formula language, modeled upon that of
arithmetic, for pure thought”
– This can be considered the ancestor of all
current computer programming languages
– Made the distinction between syntax and
semantics critical

• He invented what we today call predicate calculus (or first-order logic)

24
Using FOL to express knowledge
• One can express the knowledge of a particular domain in
first-order logic
• Example: The “kinship domain”
– Objects: people
– Properties: gender, family relationships
– Unary predicates: Male, Female
– Binary predicates: Parent, Sibling, Brother, Sister, Son,
Daughter, Father, Mother, Uncle, Aunt, Grandparent, Grandfather,
Grandmother, Husband, Wife, Spouse, Brother-in-law,
Stepmother, etc….
– Functions: MotherOf, FatherOf…

• Note: There is usually (always?) more than one way to

specify knowledge
25
Kinship domain
• Write down what we know (what we want to be in the KB)
– One’s mother is one’s female parent
  m, c Mother(m, c)  Female(m)  Parent(m, c)
  m, c TheMotherOf(c) = m  Female(m)  Parent(m, c)
– One’s husband is one’s male spouse
  w, h Husband(h, w)  Male(h)  Spouse(h, w)
– One is either male or female
  x Male(x)  Female(x)
– Parent-child relationship
  p, c Parent(p, c)  Child(c, p)
– Grandparent-grandchild relationship
  g, c Grandparent(g, c)   p Parent(g, p)  Parent(p, c)
– Etc…
• Now we can reason about family relationships. (How?)
26
Kinship domain (cont.)
Assertions (“Add this sentence to the KB”)
TELL( KB,  m, c Mother(c) = m  Female(m)  Parent(m, c) )
TELL( KB,  w, h Husband(h, w)  Male(h)  Spouse(h, w) )
TELL( KB,  x Male(x)  Female(x) )
TELL( KB, Female(Mary)  Parent(Mary, Frank)  Parent(Frank, Ann) )
– Note: TELL( KB, S1  S2 )  TELL( KB, S1) and TELL( KB, S2)
(because of and-elimination and and-introduction)

Queries (“Does the KB entail this sentence?”)

ASK( KB, Grandparent(Mary, Ann) )  True
ASK( KB,  x Child(x, Frank) )  True
– But a better answer would be  { x / Ann }
– This returns a substitution or binding

27
Simple example of inference in FOL
KB0
Bob is a buffalo Buffalo(Bob)
Pat is a pig Pig(Pat)
Buffaloes outrun pigs Buffalo(x)  Pig(y)  Outrun(x,y)

Does Bob outrun Pat? KB entails Outrun(Bob, Pat)?

S KB0

KB0 |– Buffalo(Bob)  Pig(Pat)

KB1
(And-Introduction)
KB1 |– Buffalo(Bob)  Pig(Pat)  Outrun(Bob, Pat)
KB2
(Universal Instantiation) [coming soon]
KB2 |– Outrun(Bob, Pat)
KB3
(Modus Ponens)
28
Inference in First-Order Logic
• Inference rules for propositional logic:
– Modus ponens, and-elimination, and-introduction, or-introduction,
resolution, etc.
– These are valid for FOL also

• But since these don’t deal with quantifiers and variables,

we need new rules, especially those that allow for
substitution (binding) of variables to objects
– These are called lifted inference rules

29
Substitution and variable binding
• Notation for substitution:
– SUBST ( Binding list, Sentence )
 Binding list: { var / ground term, var / ground term, … }

 “ground term” = term with no variables

– SUBST( {var/gterm}, Func (var) ) = Func (gterm)

 SUBST (, p)

– Examples:
 SUBST ( {x/Mary}, FatherOf (x) ) = FatherOf (Mary)

 SUBST ( {x/Joe, y/Lisa}, Siblings (x,y) ) = Siblings (Joe, Lisa)

– We can generalize it, in first-order logic, a substitution is a

mapping from variables to terms, i.e., constants, variables, function
(terms.))

30
Three new inference rules using SUBST(, p)
• Universal Instantiation
v  g – ground term
SUBST ({v / g },  )

• Existential Instantiation
v  k – constant that does not appear
elsewhere in the knowledge base
SUBST ({v / k },  )

• Existential Introduction
 v – variable not in 
v SUBST ({ g / v },  ) g – ground term in 
31
To Add to These Rules

32
Universal Instantiation – examples
v  g – ground term
SUBST ({v / g },  )
• x Sleepy(x)
– SUBST({x/Joe}, )
 Sleepy(Joe)

• x Mother(x)  Female(x)
– SUBST({x/Mary}, )
 Mother(Mary)  Female(Mary)
– SUBST({x/Dad}, )
 Mother(Dad)  Female(Dad)

• x, y Buffalo(x)  Pig(y)  Outrun(x,y)

– SUBST({x/Bob}, )
 y Buffalo(Bob)  Pig(y)  Outrun(Bob,y)
33
Existential Instantiation – examples
v  k – constant that does not appear
elsewhere in the knowledge base
SUBST ({v / k },  )

• x BestAction(x)
– SUBST({x/B_A}, )
 BestAction(B_A)
– “B_A” is a constant; it is not in our universe of actions

• y Likes(y, Broccoli)
– SUBST({y/Bush}, )
 Likes(Bush, Broccoli)
– “Bush” is a constant; it is not in our universe of people

34
Existential Introduction – examples
 v – variable not in 
g – ground term in 
v SUBST ({ g / v },  )

• Likes(Jim, Broccoli)
– SUBST({Jim/x}, )
 x Likes(x, Broccoli)

• x Likes(x, Broccoli)  Healthy(x)

– SUBST({Broccoli/y}, )
 y x Likes(x, y)  Healthy(x)

35
What’s our goal here?
• Formulate a search process:
– Initial state
 KB
– Operators
 Inference rules
– Goal test
 KB contains S

• What is a node?
– KB + new sentences (generated by applying the inference rules)
– In other words, the new state of the KB
• What kind of search to use?
– I.e., which node to expand next?
• How to apply inference rules?   
– Need to match the premise pattern 

36