0% found this document useful (0 votes)
91 views

Production Systems : Components of A Production System

A production system uses a set of if-then rules called production rules to operate on facts stored in working memory to solve problems. It works by matching the if portion of rules to the current working memory and executing the then portion to make changes and deduce new facts. An example is a shape identification system that uses rules about sides, angles, and other attributes to determine the type of shape.
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
91 views

Production Systems : Components of A Production System

A production system uses a set of if-then rules called production rules to operate on facts stored in working memory to solve problems. It works by matching the if portion of rules to the current working memory and executing the then portion to make changes and deduce new facts. An example is a shape identification system that uses rules about sides, angles, and other attributes to determine the type of shape.
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

Artificial Intelligence

Production Systems
Components of a production system
Working memory
Set
S t off production
d ti rules
l
Rule interpreter

Example: shapes knowledge base


Example contexts

A Production System
1. Working memory

Current state of the world

2 Set of production rules (KB)


2.

Condition action pairs (IF THEN)


Condition part matched against working memory
Action part defines the associated problem solving
step if the condition is true

3. Recognize
Recognize-Act
Act Cycle (Algorithm/Interpreter)

Initialize working memory with problem start


Repeat examining rules and taking action until the
problem is solved or nothing new can be
concluded

Artificial Intelligence

Production System

Production
Rule Base

Pattern Action

Working
Memory

State description

Rule 1 If Then

Fact 1

Fact m

Rule N If Then

Rule Interpreter Algorithm


1. If the context contains an attribute-value pair
whose attribute is the goal attribute,
then stop and output the associated value.
2. Find all productions (rules) whose antecedent
attribute-value pairs are in the context and mark
them applicable.
3. Deactivate any production which will not add a
new attribute-value pair to the context.
4. If no productions are applicable then stop and
print
i the
h message: Identification
Id ifi i unknown.
k

5. Execute the action of the lowest numbered (or


only) applicable production.
6. Reset the applicability of all productions and return
to Step 1.

Artificial Intelligence

Shape Rules
1. (IF (sides parallel)
THEN (family parallelogram))
2 (IF (sides skew)
2.
THEN (shape trapezoid))
3. (IF (sides curved)
THEN (family ellipsoid))
4 (IF (angles right)
4.
THEN (family rectangle))
5. (IF (family parallelogram)
(angles skew) (lengths equal)
THEN (shape rhombus))

Shape Rules (cont)


6. (IF (family parallelogram)(angles skew)
(lengths unequal)
THEN (shape parallelogram))
7. (IF (family rectangle) (lengths equal)
THEN (symmetrical t) (shape square))
8. (IF (family rectangle) (lengths unequal)
THEN (shape rectangle) (symmetrical t))
9. (IF (family ellipsoid) (axes unequal)
THEN (shape ellipse))
10. (IF (family ellipsoid) (axes equal)
THEN (symmetrical t) (shape circle))

Artificial Intelligence

Shape Rule Base


1. (IF (sides parallel) THEN (family parallelogram))
2. (IF (sides skew) THEN (shape trapezoid))
3. (IF (sides curved) THEN (family ellipsoid))
4. (IF (angles right) THEN (family rectangle))
5. (IF (family parallelogram) (angles skew) (lengths equal) THEN (shape
rhombus))
6. (IF (family parallelogram) (angles skew) (lengths unequal) THEN
(shape parallelogram))
7. (IF (family
y rectangle)
g (lengths
g
equal)
q
THEN (symmetrical
y
t) (shape
p
square))
8. (IF (family rectangle) (lengths unequal) THEN (shape rectangle)
(symmetrical t))
9. (IF (family ellipsoid) (axes unequal) THEN (shape ellipse))
10.(IF (family ellipsoid) (axes equal) THEN (symmetrical t) (shape circle))

Shape Contexts
1. (sides parallel)
(angles right)
(lengths equal)
2. (sides curved)
(lengths equal)
3. (angles right)
(sides parallel)
(lengths unequal)
4. (sides curved)
(color red)
(axes equal)

Artificial Intelligence

Rules for Animal Identification


1. (IF (covering hair) THEN (subclass mammal))
2. (IF (milk t) THEN (subclass mammal))
3. (IF (covering feathers) THEN (subclass bird))
4. (IF (flies t) (eggs t) THEN (subclass bird))
5. (IF (eats meat) THEN (order carnivore))
6. (IF (teeth pointed) (claws t) (eyes forward) THEN (order carnivore))
7. (IF (subclass mammal) (hoofs t) THEN (order ungulate))
8. (IF (subclass bird) (flies f) THEN (order ratite))
9. (IF (subclass mammal) (eats cud) (antlers none) THEN (order ungulate))
10. (IF (subclass mammal) (eats cud) (antlers hornlike) THEN (order ruminant))
11. (IF (subclass mammal) (eats cud) (antlers branching) THEN (order ruminant))
12. (IF (order ruminant) (antlers hornlike) THEN (toes even) (suborder pecora))
13. (IF (suborder pecora) THEN (stomachs several))
14. (IF (order ratite) (feet webbed) (wings swimming) THEN (type penguin))
15 (IF ((subclass
15.
b l mammal)
l) (order
( d carnivore)
i
) (color
( l tawny)
t
) (spots
( t dark)
d k) THEN (type
(t
cheetah))
h t h))
16. IF (subclass mammal) (order carnivore) (color tawny) (stripes black) THEN (type tiger))
17. (IF (order ungulate) (neck long) (legs long) (spots dark) THEN (type giraffe))
18. (IF (order ungulate) (stripes black) THEN (type zebra))
19. (IF (order ratite) (neck long) (legs long) (color black-white) THEN (type ostrich))
20. (IF (subclass bird) (flies t) THEN (type albatross))

Contexts for Animal Rule Base


1. (covering feathers) (swims t) (wings
swimming) (feet webbed) (files f)
2 (covering
2.
(
i
hair)
h i ) (eats
(
cud)
d) (legs
(l
long)
l
)
(color tawny) (spots dark)
(antlers none) (neck long)
3. (covering hair) (eats meat) (color
tawny) (teeth pointed) (claws t)
(eyes forward)
4. (eats cud) (milk t) (toes even)
(antlers hornlike)
5. (covering hair) (eats meat)
(color tawny) (spots dark)

You might also like