Logic Agents and Propositional

Logic

CHAPTER 7
HASSAN KHOSRAVI
SPRING2011
 Knowledge-Based Agents

 KB = knowledge base
 A set of sentences or facts
 e.g., a set of statements in a logic language

 Inference
 Deriving new sentences from old
 e.g., using a set of logical statements to infer new ones

 A simple model for reasoning

 Agent is told or perceives new evidence
 E.g., A is true
 Agent then infers new facts to add to the KB
 E.g., KB = { A -> (B OR C) }, then given A and not C we can infer that B is true
 B is now added to the KB even though it was not explicitly asserted, i.e., the
agent inferred B
 Wumpus World

 Environment
 Cave of 4×4

 Agent enters in [1,1]

 16 rooms
 Wumpus: A deadly beast who kills anyone
entering his room.
 Pits: Bottomless pits that will trap you
forever.
 Gold
 Wumpus World

 Agents Sensors:
 Stench next to Wumpus
 Breeze next to pit
 Glitter in square with gold
 Bump when agent moves into a wall
 Scream from wumpus when killed

 Agents actions
 Agent can move forward, turn left or
turn right
 Shoot, one shot
 Wumpus World

 Performance measure
 +1000 for picking up gold

 -1000 got falling into pit

 -1 for each move

 -10 for using arrow

 Reasoning in the Wumpus World

 Agent has initial ignorance about the configuration

 Agent knows his/her initial location
 Agent knows the rules of the environment

 Goal is to explore environment, make inferences

(reasoning) to try to find the gold.

 Random instantiations of this problem used to test agent

reasoning and decision algorithms

(applications? “intelligent agents” in computer games)

 Exploring the Wumpus World

[1,1] The KB initially contains the rules of the environment.

The first percept is [none, none,none,none,none],

move to safe cell e.g. 2,1

 Exploring the Wumpus World

[2,1] = breeze

indicates that there is a pit in [2,2] or [3,1],

return to [1,1] to try next safe cell

 Exploring the Wumpus World

[1,2] Stench in cell which means that wumpus is in [1,3] or [2,2]

YET … not in [1,1]
YET … not in [2,2] or stench would have been detected in [2,1]
(this is relatively sophisticated reasoning!)
 Exploring the Wumpus World

[1,2] Stench in cell which means that wumpus is in [1,3] or [2,2]

YET … not in [1,1]
YET … not in [2,2] or stench would have been detected in [2,1]
(this is relatively sophisticated reasoning!)
THUS … wumpus is in [1,3]
THUS [2,2] is safe because of lack of breeze in [1,2]
THUS pit in [1,3] (again a clever inference)
move to next safe cell [2,2]
 Exploring the Wumpus World

[2,2] move to [2,3]

[2,3] detect glitter , smell, breeze

THUS pick up gold
THUS pit in [3,3] or [2,4]
 What our example has shown us

 Can represent general knowledge about an environment by a

set of rules and facts

 Can gather evidence and then infer new facts by combining

evidence with the rules

 The conclusions are guaranteed to be correct if

 The evidence is correct
 The rules are correct
 The inference procedure is correct
-> logical reasoning

 The inference may be quite complex

 E.g., evidence at different times, combined with different rules, etc
 What is a Logic?

 A formal language
 KB = set of sentences

 Syntax
 what sentences are legal (well-formed)
 E.g., arithmetic
 X+2 >= y is a wf sentence, +x2y is not a wf sentence

 Semantics
 loose meaning: the interpretation of each sentence
 More precisely:
 Defines the truth of each sentence wrt to each possible world
 e.g,
 X+2 = y is true in a world where x=7 and y =9
 X+2 = y is false in a world where x=7 and y =1

 Note: standard logic – each sentence is T of F wrt eachworld

 Fuzzy logic – allows for degrees of truth.
 Models and possible worlds

 Logicians typically think in terms of models, which are formally

structured worlds with respect to which truth can be evaluated.

 m is a model of a sentence  if  is true in m

 M() is the set of all models of 

 Possible worlds ~ models

 Possible worlds: potentially real environments
 Models: mathematical abstractions that establish the truth or falsity of every
sentence

 Example:
 x + y = 4, where x = #men, y = #women
 Possible models = all possible assignments of integers to x and y
 Entailment

 One sentence follows logically from another

 |= b

 entails sentence b if and only if b is true in all

worlds where  is true.

e.g., x+y=4 |= 4=x+y

 Entailment is a relationship between sentences that

is based on semantics.
 Entailment in the wumpus world

 Consider possible models for KB assuming only pits and a reduced Wumpus
world

 Situation after detecting nothing in [1,1], moving right, detecting breeze in

[2,1]
 Wumpus models

All possible models in this reduced Wumpus world.

 Wumpus models

 KB = all possible wumpus-worlds consistent

with the observations and the “physics” of the
Wumpus world.
 Inferring conclusions

 Consider 2 possible conclusions given a KB

 α1 = "[1,2] is safe"
 α2 = "[2,2] is safe“

 One possible inference procedure

 Start with KB
 Model-checking
 Check if KB ╞  by checking if in all possible models where KB is
true that  is also true

 Comments:
 Model-checking enumerates all possible worlds
 Only works on finite domains, will suffer from exponential growth
of possible models
 Wumpus models

α1 = "[1,2] is safe", KB ╞ α1, proved by model checking

 Wumpus models

α2 = "[2,2] is safe", KB ╞ α2
There are some models entailed by KB where 2 is
false
 Logical inference

 The notion of entailment can be used for logic inference.

 Model checking (see wumpus example): enumerate all possible
models and check whether  is true.

 If an algorithm only derives entailed sentences it is called

sound or truth preserving.
 Otherwise it just makes things up.
i is sound if whenever KB |-i  it is also true that KB|= 
 E.g., model-checking is sound

 Completeness : the algorithm can derive any sentence

that is entailed.
i is complete if whenever KB |=  it is also true that KB|-i 
 Schematic perspective

If KB is true in the real world, then any sentence  derived

from KB by a sound inference procedure is also true in the
real world.
 Propositional logic: Syntax

 Propositional logic is the simplest logic – illustrates basic ideas

 Atomic sentences = single proposition symbols

 E.g., P, Q, R
 Special cases: True = always true, False = always false

 Complex sentences:

 If S is a sentence, S is a sentence (negation)

 If S1 and S2 are sentences, S1  S2 is a sentence (conjunction)

 If S1 and S2 are sentences, S1  S2 is a sentence (disjunction)

 If S1 and S2 are sentences, S1  S2 is a sentence (implication)

 If S1 and S2 are sentences, S1  S2 is a sentence (biconditional)

 Propositional logic: Semantics

Each model/world specifies true or false for each proposition symbol

E.g. P1,2 P2,2 P3,1
false true false
With these symbols, 8 possible models, can be enumerated automatically.

Rules for evaluating truth with respect to a model m:

S is true iff S is false

S1  S2 is true iff S1 is true and S2 is true

S1  S2 is true iff S1is true or S2 is true

S1  S2 is true iff S1 is false or S2 is true

i.e., is false iff S1 is true and S2 is false

S1  S2 is true iff S1S2 is true andS2S1 is true

Simple recursive process evaluates an arbitrary sentence, e.g.,

P1,2  (P2,2  P3,1) = true  (true  false) = true  true = true

Truth tables for connectives
 Truth tables for connectives

Implication is always true

when the premise is false

Why? P=>Q means “if P is true then I am claiming that Q is true,

otherwise no claim”
Only way for this to be false is if P is true and Q is false
 Wumpus world sentences
Let Pi,j be true if there is a pit in [i, j].
Let Bi,j be true if there is a breeze in [i, j].
start:  P1,1
 B1,1
B2,1

 "Pits cause breezes in adjacent squares"

B1,1  (P1,2  P2,1)
B2,1  (P1,1  P2,2  P3,1)

 KB can be expressed as the conjunction of all of these sentences

 Note that these sentences are rather long-winded!

 E.g., breese “rule” must be stated explicitly for each square
 First-order logic will allow us to define more general relations (later)
Truth tables for the Wumpus KB
 Inference by enumeration

 We want to see if  is entailed by KB

 Enumeration of all models is sound and complete.

 But…for n symbols, time complexity is O(2n)...

 We need a more efficient way to do inference

 But worst-case complexity will remain exponential for
propositional logic
 Logical equivalence

 To manipulate logical sentences we need some rewrite rules.

 Two sentences are logically equivalent iff they are true in same models: α ≡ ß iff α╞ β and
β╞ α
 Modus Ponens

 And-Elimination

 Bi-conditional Elimination
 Validity and satisfiability

A sentence is valid if it is true in all models,

e.g., True, A A, A  A, (A  (A  B))  B
(tautologies)
Validity is connected to inference via the Deduction
Theorem:
KB ╞ α if and only if (KB  α) is valid

A sentence is satisfiable if it is true in some model

e.g., A B, C
(determining satisfiability of sentences is NP-complete)
A sentence is unsatisfiable if it is false in all models
e.g., AA
Satisfiability is connected to inference via the following:
KB ╞ α if and only if (KB α) is unsatisfiable
(there is no model for which KB=true and  is false)
(aka proof by contradiction: assume  to be false and this leads to
contraditions in KB)
 Proof methods

 Proof methods divide into (roughly) two kinds:

Application of inference rules:

Legitimate (sound) generation of new sentences from old.
 Resolution
 Forward & Backward chaining

Model checking
Searching through truth assignments.
 Improved backtracking: Davis--Putnam-Logemann-Loveland (DPLL)
 Heuristic search in model space: Walksat.
 Normal Form

We want to prove: KB | 
equivalent to : KB   unsatifiable

We first rewrite KB   into conjunctive normal form (CNF).

literals
A “conjunction of disjunctions”
(A  B)  (B  C  D)

Clause Clause

• Any KB can be converted into CNF

• k-CNF: exactly k literals per clause
 Example: Conversion to CNF

B1,1  (P1,2  P2,1)

1. Eliminate , replacing α  β with (α  β)(β  α).

(B1,1  (P1,2  P2,1))  ((P1,2  P2,1)  B1,1)

2. Eliminate , replacing α  β with α β.

(B1,1  P1,2  P2,1)  ((P1,2  P2,1)  B1,1)

3. Move  inwards using de Morgan's rules and double-negation:

(B1,1  P1,2  P2,1)  ((P1,2  P2,1)  B1,1)

4. Apply distributive law ( over ) and flatten:

(B1,1  P1,2  P2,1)  (P1,2  B1,1)  (P2,1  B1,1)
 Resolution Inference Rule for CNF
(A  B  C ) “If A or B or C is true, but not A, then B or C
( A ) must be true.”

 (B  C )
“If A is false then B or C must be true,
(A  B  C ) or if A is true then D or E must be true,
hence since A is either true or false, B or C
( A  D  E )
or D or E must be true.”

 (B  C  D  E )

(A  B )
( A  B ) Simplification

 (B  B )  B
 Resolution Algorithm

• The resolution algorithm tries to prove:KB |  equivalent to

KB   unsatisfiable

• Generate all new sentences from KB and the query.

• One of two things can happen:

1. We find P  P which is unsatisfiable,

i.e. we can entail the query.

2. We find no contradiction: there is a model that satisfies the

Sentence (non-trivial) and hence we cannot entail the query.
KB  
 Resolution example

 KB = (B1,1  (P1,2 P2,1))  B1,1

 α = P1,2
KB  

True
False in
all worlds
 Horn Clauses

• Resolution in general can be exponential in space and time.

• If we can reduce all clauses to “Horn clauses” resolution is linear in space and time

A clause with at most 1 positive literal.

e.g. A  B  C

• Every Horn clause can be rewritten as an implication with

a conjunction of positive literals in the premises and a single
positive literal as a conclusion.
e.g. B  C  A

• 1 positive literal: definite clause

• 0 positive literals: Fact or integrity constraint:

e.g. (A  B )  (A  B  False )
Forward-chaining pseudocode
 Forward chaining: graph representation

 Idea: fire any rule whose premises are satisfied in the

KB,
 add its conclusion to the KB, until query is found

AND gate

OR gate

• Forward chaining is sound and complete for Horn KB

 Forward chaining example

“OR” Gate

“AND” gate
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
Forward chaining example
 Forward chaining

 FC is data-driven, automatic, unconscious

processing,
 e.g., object recognition, routine decisions

 May do lots of work that is irrelevant to the goal

 Backward chaining

Idea: work backwards from the query q

 check if q is known already, or
 prove by BC all premises of some rule concluding q
 Hence BC maintains a stack of sub-goals that need to be proved to get to q.
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example

we need P to prove
L and L to prove P.
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
Backward chaining example
 Backward chaining

 BC is goal-driven, appropriate for problem-solving,

 e.g., Where are my keys? How do I get into a PhD program?

 Complexity of BC can be much less than linear in size of KB

Avoid loops: check if new sub-goal is already on the goal stack

Avoid repeated work: check if new sub-goal

1. has already been proved true, or
2. has already failed

Like FC, is linear and is also sound and complete (for Horn KB)
 Model Checking

Two families of efficient algorithms:

 Complete backtracking search algorithms: DPLL

algorithm

 Incomplete local search algorithms

 WalkSAT algorithm
 Satisfiability problems

 Consider a CNF sentence, e.g.,

(D  B  C)  (B  A  C)  (C  B  E) 
(E  D  B)  (B  E  C)

Satisfiability: Is there a model consistent with this sentence?

[A  B]  [¬B  ¬C]  [A  C]  [¬D]  [¬D  ¬A]

 The WalkSAT algorithm

 Incomplete, local search algorithm

 Begin with a random assignment of values to symbols
 Each iteration: pick an unsatisfied clause
 Flip the symbol that maximizes number of satisfied clauses, OR
 Flip a symbol in the clause randomly
 Trades-off greediness and randomness
 Many variations of this idea
 If it returns failure (after some number of tries) we cannot
tell whether the sentence is unsatisfiable or whether we
have not searched long enough
 If max-flips = infinity, and sentence is unsatisfiable, algorithm never
terminates!

 Typically most useful when we expect a solution to exist

Pseudocode for WalkSAT
 Hard satisfiability problems

 Consider random 3-CNF sentences. e.g.,

(D  B  C)  (B  A  C)  (C  B  E) 
(E  D  B)  (B  E  C)

m = number of clauses (5)

n = number of symbols (5)

 Underconstrained problems:
 Relatively few clauses constraining the variables
 Tend to be easy
 16 of 32 possible assignments above are solutions
 (so 2 random guesses will work on average)
 Hard satisfiability problems

 What makes a problem hard?

 Increase the number of clauses while keeping the number of
symbols fixed
 Problem is more constrained, fewer solutions

 Investigate experimentally….
P(satisfiable) for random 3-CNF sentences, n
= 50
 Run-time for DPLL and WalkSAT

 Median runtime for 100 satisfiable random 3-CNF sentences, n = 50

Inference-based agents in the wumpus world

A wumpus-world agent using propositional logic:

P1,1 (no pit in square [1,1])

W1,1 (no Wumpus in square [1,1])
Bx,y  (Px,y+1  Px,y-1  Px+1,y  Px-1,y) (Breeze next to Pit)
Sx,y  (Wx,y+1  Wx,y-1  Wx+1,y  Wx-1,y) (stench next to Wumpus)
W1,1  W1,2  …  W4,4 (at least 1 Wumpus)
W1,1  W1,2 (at most 1 Wumpus)
W1,1  W8,9
…

 64 distinct proposition symbols, 155 sentences

Limited expressiveness of propositional logic

 KB contains "physics" sentences for every single square

 For every time t and every location [x,y],

Lx,y  FacingRightt  Forwardt  Lx+1,y

 Rapid proliferation of clauses.

First order logic is designed to deal with this through the

introduction of variables.
 Summary

 Logical agents apply inference to a knowledge base to derive new

information and make decisions

 Basic concepts of logic:

 syntax: formal structure of sentences
 semantics: truth of sentences wrt models
 entailment: necessary truth of one sentence given another
 inference: deriving sentences from other sentences
 soundness: derivations produce only entailed sentences
 completeness: derivations can produce all entailed sentences

 Resolution is complete for propositional logic

 Forward, backward chaining are linear-time, complete for Horn clauses

 Propositional logic lacks expressive power