Artificial Intelligence—15-381

Unification and Resolution in FOL

Jaime Carbonell
11-February-2003
OUTLINE:
 Q/A by matching (partial unification)
 Full term unification in FOL

 FOL in clause form (with Skolemization)

 FOL Robinson Resolution

 Uses and limitations of Resolution

 Answering Questions by Matching

Data Base of Possible Queries

Ground Instances
Person(x)
Person(John) Likes(Mary,x)
Person(Mary) Likes(x,x)
Likes(x,y)
Dog(Fido) Not(Likes(z, John))
Likes(John,Mary) Owners(Fido,x,y)
Not(Likes (Mary, John)) Simple Matching
Likes(Fido, Mary)
Token match on predicates
Likes(Fido, John)
Variables only for query
Likes(Fido, Fido)
arguments const-const or const-
Owners(Fido, John, Mary)
variable match
Zero, one, or multiple answers
 Full Unification
(variables on both sides)
Variable-Variable Matching
P(x) P(y)
Q(x,x) Q(y,z)
P(f(x),z) P(y, Fido)
Unifiers: Variable Substitutions
P(x) P(y) {y/x}
Q(x,x) Q(y,z) {y/x, z/x}
P(f(x),z) P(y,Fido) {y/f(x), z/Fido}
Consistent Variable Assignments
P(Mary,John) P(y,y) #
R(x,y,y) R(y,y,z) {y\z, x\y}
W(P(x),y,z) W(Q(x),y,Fido) #
Advantages of Full Unification
 Query and data => both fully allow variables
 Permits full FOL Resolution (next)
 Unification

Q(x)
P(y)  FAIL
P(x)
P(y)  x/y
P(Marcus)
P(y)  Marcus/y
P(Marcus)
P(Julius)  FAIL
P(x,x)
P(y,y)  (y/x)
P(y,z)  (z/y , y/x)
 Finding General Substitutions

Given:
Hate(x,y)
Hate(Marcus,z)

We Could Produce:
(Marcus/x, z/y)
(Marcus/x, y/z)
(Marcus/x, Caesar/y, Caesar/z)
(Marcus/x, Polonius/y, Polonius/z)
 Algorithm: Unify(L1,L2)
1. If L1 or L2 is a variable or 4. Set SUBST to NIL.
constant, then:
a) If L1 and L2 are identical, then 5. For i  1 to number of arguments in
return NIL.
b) Else if L1 is a variable, then if L1:
L1 occurs in L2 then return a) Call Unify with the ith argument
FAIL, else return {(L2/L1)}. of L1 and the ith argument of L2,
c) Else if L2 is a variable, then if putting result in S.
L2 occurs in L1 then return b) If S = FAIL then return FAIL.
FAIL, else return {(L1/L2)}. c) If S is not equal to NIL then:
d) Else return FAIL. i. Apply S to the remainder of both
L1 and L2.
2. If the initial predicate symbols in ii. SUBST := APPEND(S, SUBST).
L1 and L2 are not identical, then
return FAIL.
6. Return SUBST.
3. If L1 and L2 have a different
number of arguments, then return
FAIL
 A Predicate Logic Example
Marcus was a man. man(Marcus)
Marcus was a Pompeian. Pompeian(Marcus)
All Pompeians were Romans. ∀x: Pompeian(x)→Roman(x)
Caesar was a ruler. ruler(Caesar)
All Romans were either loyal to Caesar or hated him.
∀x: Roman(x) → loyalto(x, Caesar) V hate(x,Caesar)
Everyone is loyal to someone. ∀x: ∃y: loyalto(x,y)
People only try to assassinate rulers they aren't loyal to.
∀x:∀y:person(x)Λ ruler(y) Λ tryassassinate(x,y)→¬ loyalto(x,y)
10. Marcus tried to assassinate Caesar.
tryassassinate(Marcus, Caesar)
All men are people. ∀x: man(x) → person(x)
 Conversion to Clause Form

Problem:
∀x: [Roman(x)Λ know(x,Marcus)]→ [hate(x,Caesar) V (∀y:∃z:
hate(y,z) → thinkcrazy(x,y))]
Solution:
 Flatten
 Separate out quantifiers
Conjunctive Normal Form: ¬Roman(x) v ¬know(x,Marcus) v
hate(x,Caesar) v ¬ hate(y,z) v thinkcrazy(x,z)
Clause Form
 Conjunctive normal form
 No instances of Λ
 Algorithm: Convert to Clause Form
Eliminate →, using: a → b= ¬ a v b.
Reduce the scope of each ¬ to a single term, using:
∀ ¬ (¬ p) = p
• deMorgan's laws: ¬(a Λ b) = ¬ a V ¬ b
¬(a V b) = ¬ a Λ ¬ b
∀ ¬ ∀x P(x) = ∃x ¬ P(x)
∀ ¬ ∃x P(x) = ∀x ¬ P(x)
Standardize variables.
Move all quantifiers to the left of the formula without changing their
relative order.
Eliminate existential quantifiers by inserting Skolem functions.
Drop the prefix.
Convert the expression into a conjunction of disjuncts, using
associativity and distributivity.
Create a separate clause for each conjunct.
Standardize apart the variables in the set of clauses generated in step 8,
using the fact that: (∀x: P(x) Λ Q(x)) = ∀x: P(x) Λ ∀x: Q(x)
 Skolem Functions in FOL
Objective
 Want all variables universally quantified
 Notational variant of FOL w/o existentials
 Retain implicitly full FOL expressiveness
Skolem’s Theorem
Every existentially quantified variable can be replaced by a unique
Skolem function whose arguments are all the universally quantified
variables on which the existential depends, without changing FOL.
Examples
 “Everybody likes something”
∀(x) ∃ (y) [Person(x) & Likes(x,y)]
∀(x) [Person(x) & Likes(x, S1(x))]
Where S1(x) = “that which x likes”
 “Every philosopher writes at least one book”
∀(x) ∃(y)[Philosopher(x) & Book(y)) => Write(x,y)]
∀(x)[(Philosopher(x) & Book(S2(x))) => Write(x,S2(x))]
 Examples of Conversion to Clause Form
Example:
∀x: [Roman(x) Λ know(x, Marcus)] →[hate(x,Caesar) V (∀y: ∃z:
hate(y,z) → thinkcrazy(x,y))]
 Eliminate →
∀x: ¬[Roman(x) Λ know(x, Marcus)] V [hate(x,Caesar) V (∀y:
¬∃z: hate(y,z) V thinkcrazy(x,y))]
 Reduce scope of ¬.
∀x: [ ¬Roman(x) V ¬ know(x, Marcus)] V [hate(x,Caesar) V (∀y:
∀z: ¬hate(y,z) V thinkcrazy(x,y))]
 “Standardize” variables:
∀x: P(x) V ∀x: Q(x) converts to ∀x: P(x) V ∀y: Q(y)
 Move quantifiers. ∀x: ∀y: ∀z: [¬Roman(x) V ¬ know(x,
Marcus)] V [hate(x,Caesar) V (¬hate(y,z) V thinkcrazy(x,y))]
 Examples of Conversion to Clause Form

 Eliminate existential quantifiers.

∃y: President(y) will be converted to President(S1)
∀x: ∃y: father-of(y,x) will be converted to ∀x: father-of(S2(x),x))
 Drop the prefix.
[ ¬Roman(x) ¬ know(x,Marcus)] V [hate(x, Caesar) V (¬
hate(y,z) V thinkcrazy(x,y))]
 Convert to a conjunction of disjuncts.
¬ Roman(x) V ¬ know(x,Marcus) V hate(x,Caesar) V ¬ hate(y,z)
V thinkcrazy(x,y)
 The Basis of Resolution and
Herbrand's Theorem

Given:
winter V summer
¬ winter V cold
We can conclude:
summer v cold

Herbrand's Theorem:
To show that a set of clauses S is unsatisfiable, it is
necessary to consider only interpretations over a particular
set, called the Herbrand universe of S. A set of clauses S is
unsatisfiable if and only if a finite subset of ground
instances (in which all bound variables have had a value
substituted for them) of S is unsatsifable.
 Algorithm: Propositional Resolution
Convert all the propositions of F to clause form.
Negate P and convert the result to clause form. Add it to
the set of clauses obtained in step 1.
Repeat until either a contradiction is found or no progress
can be made:
a) Select two clauses. Call these the parent clauses.
b) Resolve them together. The resolvent will be the
disjunction of all of the literals of both of the parent clauses
with the following exception: If there are any pairs of
literals L and ¬ L such that one of the parent clauses contains L
and the other contains ¬L, then select one such pair and eliminate
both L and ¬L from the resolvent.
c) If the resolvent is the empty clause, then a contradiction has been
found. If it is not, then add it to the set of clauses available to the
procedure.
 A Few Facts in Propositional Logic

Given Axioms Clause Form

P P (1)

(P Λ Q) → R ¬P V ¬Q V R (2)

(S V T) → Q ¬S V Q (3)

¬T V Q (4)

T T (5)
Resolution in Propositional Logic

¬P V ¬Q V R ¬R

¬P V ¬Q P

¬T V Q ¬Q

¬T T
 Algorithm: Resolution
Convert all the propositions of F to clause form.
Negate P and convert the result to clause form. Add it to the
set of clauses obtained in 1.
Repeat until either a contradiction is found, no progress can be
made, or a predetermined amount of effort has been expended.
a) Select two clauses. Call these the parent clauses.
b) Resolve them together. The resolvent will be the disjunction of all
the literals of both parent clauses with appropriate substitutions
performed and with the following exception: If there is one pair of
literals T1 and ¬ T2 such that one of the parent clauses contains T1
and the other contains ¬ T2 and if T1 and T2 are unifiable, then
neither T1 nor ¬ T2 should appear in the resolvent. If there is more
than one pair of complementary literals, only one pair should be
omitted from the resolvent.
c) If the resolvent is the empty clause, then a contradiction has been
found. If it is not, then add it to the set of clauses available to the
procedure.
 A Resolution Proof

Axioms in clause form:

1. man(Marcus)
2. Pompeian(Marcus)
3. ¬ Pompeian(x1) v Roman(x1)
4. Ruler(Caesar)
5. ¬ Roman(x2) v loyalto(x2, Caesar) v hate(x2, Caesar)
6. loyalto(x3, f1(x3))
7. ¬ man(x4) v ¬ ruler(y1) v ¬ tryassassinate(x4, y1) v
loyalto (x4, y1)
8. tryassassinate(Marcus, Caesar)
 Resolution Proof cont.
Prove: hate(Marcus, Caesar) ¬hate(Marcus, Caesar) 5

Marcus/x2
3 ¬Roman(Marcus) V loyalto(Marcus,Caesar)

Marcus/x1
¬Pompeian(Marcus) V loyalto(Marcus,Caesar) 2

7 loyalto(Marcus,Caesar)

Marcus/x4, Caesar/y1

1 ¬man(Marcus) V ¬ ruler(Caesar) V ¬ tryassassinate(Marcus, Caesar)

¬ ruler(Caesar) V ¬ tryassassinate(Marcus, Caesar) 4

¬ tryassassinate(Marcus, Caesar) 8
 An Unsuccessful Attempt at Resolution
Prove: loyalto(Marcus, Caesar) ¬loyalto(Marcus, Caesar) 5

Marcus/x2

3 ¬Roman(Marcus) V hate(Marcus,Caesar)
Marcus/x1
¬Pompeian(Marcus) V hate(Marcus,Caesar) 2

hate(Marcus,Caesar)
(a)
hate(Marcus,Caesar) 10

Marcus/x6, Caesar/y3
persecute(Caesar, Marcus) 9
Marcus/x5, Caesar/y2
hate(Marcus,Caesar)
:
: (b)
 Using Resolution with Equality and Reduce
Axioms in clause form:
man(Marcus)
Pompeian(Marcus)
Born(Marcus, 40)
¬ man(x1) V mortal(x1)
¬ Pompeian(x2) V died(x2,79)
erupted(volcano, 79)
¬ mortal(x3) V ¬ born(x3, t1) V ¬gt(t2—t1, 150) V dead(x3, t2)
Now=2002
¬ alive(x4, t3) V ¬dead (x4, t3)
¬ dead(x5, t4) V alive (x5, t4)
¬ died (x6, t5) V ¬ gt(x6, t5) V dead(x6, t6)

Prove: ¬alive(Marcus, now)

 Issues with Resolution

Requires full formal representation in FOL (for

conversion to clause form)
Resolution defines a search space (which clauses
will be resolved against which others define the
operators in the space)  search method required
Worst case: resolution is exponential in the number
of clauses to resolve. Actual: exponential in
average resolvable set (= branching factor)
Can we define heuristics to guide search for BestFS,
or A* or B*? (Not in the general case)