Rule Based Systems- Introduction

 Well-formed formula in propositional logic or FOL

represent assertion knowledge
 Such wffs are divided into two categories
 Rules: Assertion given in implication form
 Facts: assertions that represent domain specific
knowledge
 In logic we represent knowledge in a declarative and
static way – as some facts and rules that are true
 Rules in logic say what is TRUE given some condition
 Rule-based systems are based on rules that say what to
do, given various conditions
 A special interpreter controls when rules are invoked
 Simple examples are very similar in rules in logic
 However, in rule based systems we consider
 Other kinds of actions (apart from adding facts)
 Degree of certainty associated with facts
 Various different control schemes (not necessarily
related to idea of logical proof)
Example Rules
Modus Ponens and Rule-
Chaining
 A system whose KB is represented as a set of rules and
facts is called a Rule Based System
 A rule-based system consists of a collection of IF-
THEN rules, a collections of facts and some interpreter
controlling the application of the rules, given the facts
 Rules are represented in the following form:
 IF <antecedent> THEN <consequent>
 When the antecedent part is NULL then Rule becomes
a fact
 Rules are normally represented as HORN clause
HC
 Horn clauses can have at most one non-negative clause
 Triggered and fired rules
 A rule is triggered when all the antecedents evaluate to
true
 A rule is fired when the action stated in the
consequent part of the inference related to the
consequent part is inferred
Rule-based system architecture
Inference Mechanism
 A machine that implements strategies to utilise the KB
and derive new conclusion from it
 Automata of IM (is the study of mathematical object)
 The execute state fires the rules once all its
antecedents match
 essentially, the function of the execute state can be
thought of as searching a path to the goal in a search
space
Example
Backtracking
Conflict resolution
 The objective is to decide which of the triggered rules
in a particular stage should be fired
 Can use various strategies
 FCFC (first come first serve) (rule ordering
 Specificity ordering
 Fire all
 Heuristic measure (distance from the goal)
example
Conflict resolution strategies
 Refractoriness: Rules once fired will not be fired later
 Meta-rules: rules about rules embedded within the
inference machine that provide the information about
which of the rules apply under what conditions.
 90% of computing time is spend in the match phase,
hence minimizing the same is necessary
Searching the space
 Start from the given facts try to arrive at the goal [data-
driven]
 Start from the goal and try to prove the goal using the
given facts [goal-driven]
 In the case of rule-based system how is the search
space defined?
 Example???
 Reasoning mechanism in rule-based systems
Forward chaining mechanism
(data-driven)
 Starting from the start state applying rules one by one
to arrive at the goial state
 Also known as data-driven search
 Forward chaining may lead search to a dead end. In
such cases backtracking is necessary [give example]
 Backtracking strategies can be
 Chronological
 intelligent
 In Fc system facts are held ina WM
 Copndition action rules represent actions to take
when speciofied facts occur in WM
 Typically the actions involve adding or deleting facts
from WM
 control
Forward Chaining
 Control cycle called recognise –act cycle
 Repeat
 Find all rules which have satisfied conditions given facts
in working memory
 Choose one, using conmflict resolution strategy
 Perform action in conclusion, probably modifying WM
 Until no rules can fire, or halt symbol added to WM
Example
 Simple “fire” example from earlier
 Working memory initially contains
 Alarm-beeps
 Hot
 Following the algorithm: first cycle
 Find all rules with satisfied (match)
 Condition: R2
 Choose one R2
 Perform action: ADD smokes
 Working memory contains
 Alarm-beeps, hot, smokey
 Continue next cycle and write yourself
Forward Chaining Applications
 Forward chaining systems have been used as:
 A model of human reasoning
 Basis for expert systems: various expert system shells
based on this model, such as CLIPS
 Practical forward chaining systems supportr pattern
matching
 Example CLIPS rule:
 (default fire-alarm
 (temperature? R1 hot
 environment? R2 smoky)
 =>
 (assert (fire-in? r1))
Architecture
Backward chaining
 Some rules/facts may be processed differently, using
backward chaining interpreter
 This allows rather more focused style of reasoning
(forward chaining may result in a lot of irrelevent
conclusions added in working memory/)
 Start with possible hypothesis, Should I switch the
sprinklers on?
 Set this as a goal to prove
 Similar to Prolog which uses a backward chaining style
of reasoning
Backward chaining
 Basic algorithm
 To prove goal G:
 If G is in the initial facts, it is proven
 Otherwise, find a rule which can be used to conclude G
and try to prove each of that rule’s condition
example
Backward chaining example
 Should we switch_on the sprinklers? Set as a goal:
 G1: switch_on_sprinklers
 Is it the initial facts? No, is there a rule which adds this
as a conclusion? Yes, re
 Set condition of r3 as new goal to prove
 G2: fire
 Is it in initial facts: No, Rule? Yes, r1
 Set conditions as new goals: G3: hot, G4 smokey
 continue
Examples of RBS
 Available expert system tools
 CLIPS
 NEXPERT
 Language facilitating development of Rule Based
Systems by incorporating within in structure certain
advanced storage techniques
 PROLOG
Expert systems
 Systems acting in a particular domain and behaving
like human expert therein
 Normally expert systems are applied to problems for
which no algorithmic solution exist
 Guided by domain sp[ecific rules and should be able to
explain irts actions
ES applications
 Rule Based Systems have been widely used in expert
systems
 E.g. medical systems, where start with set of
hypothesis on possible disease and try to prove each
one, asking addittional nquestions of using when fact
is known
 MYCIN pioneering
Vehicle diagnostic system
R1: If gas_in_engine and turns_over, then problem (Spark_plugs)
R2: If not(turns_over) and not (lights_on) then problem (battery)
R3: If not (turns_over) and lights_on, then problem (starter)
R4: If gas_in_turn and gas_in_carb then gas_in_engine

Facts
Known: gas_in_tank Solve this using FC and BC
Known: gas_in_carb Supose the goal is to prove problem
Known: gas_in_engine (battery) with the same set of initial
Known: turns_over facts. Draw the search space based
X = spark_plugs (by R1) on the rules fired.
What is Frame?
 A frame is a prototype of a concept
 Denoting the attributes of the concept
 The class of objects or concepts to which the concept in
question..
 And some more things
 An instance of a frame is a representation of a
SPECIFIC object
 Instantiation of a frame
Example
 Fido is a dog
 The concept dog
 Fido is a particular dog
 How can we represent the concept/notion of the
creature “dog”?
 Frame consists of number of slots
Economy of represenation
Object attribute value
 An instance of a frame denotes an object
 A frame denotes a class to which the object belongs
 The slots denote the attributes of the object
 The slots can be of different types
 The slots are instantiated with values
slots
 Slots denote attributes
 Attributes are typed
 Some values are defined in the frame definition
 Some values are defined in the instances only
 Instantiated slots define facts may be used to answer
queries
What are there in a slot
 Values
 Types
 Constraints over possible values
 Predicates
 Comapre with object oriented systems
 Slots can be structured frames within slots pointers
frames
Inferencing in a Frame
 Generic values
 Default value
 Multiple inheritence