Download as pptx, pdf, or txt
Download as pptx, pdf, or txt
You are on page 1of 66

Chapter 4

Knowledge Application Systems:


Systems that Utilize Knowledge

1
Artificial Intelligence

• Artificial Intelligence (AI) is the area of computer


science that deals with the design and
development of computer systems that exhibit
human-like cognitive capabilities..
• Modern AI systems recognized that intelligence
and knowledge are tightly intertwined
• Knowledge is associated with the cognitive symbols
we manipulate,
 while human intelligence refers to our ability to learn and
communicate in order to solve problems
Ch
2
AI ...

• One of the popular area of AI application is knowledge- based


systems, which we refer to here as knowledge application
systems..
• In this chapter, we discuss two popular knowledge
application systems:
 rule-based expert systems and

 case-based reasoning

Ch
3
Rule Based Expert System

• Traditionally, the development of knowledge-


based systems had been based on the use of
rules or models to represent the domain
knowledge.
• Requires collaboration of a subject matter
expert with a knowledge engineer,
• knowledge engineer is responsible for the
elicitation and representation of the expert’s
knowledge.
Rule Based …

• Developing knowledge application systems requires


eliciting knowledge from the expert and represent in a
form usable by computers, called knowledge
engineering.
• The knowledge is represented more commonly in a set
of heuristics, or rules-of-thumb.
• Experts develop these rules-of-thumb over years of
practical experience at solving problems.
• This rules-of-thumb must be represented as production
rules or IF-THEN statements.
Rule Based …

• Rules are the most commonly used knowledge


representation paradigm
• The IF portion is the condition which tests the
truth-value of a set of assertions.
• If the statement is true, the THEN part of the rule
(also action, conclusion, or consequence) is
also inferred as a fact.
• Other paradigms to represent knowledge include
frames, predicates, associative networks, and
objects.
Simple Example of ES

• IF patient has pain • IF patient has pain


THEN prescribe pain AND patient is over
killers (priority 10) 60
• IF patient has chest AND patient has a
pain history of heart
THEN treat for heart conditions
disease (priority 100) THEN take to
emergency room

Ch
Rule based ...

• Rule based expert systems were among the first truly


successful forms of Artificial Intelligence software.
• An expert system is divided into two sub-systems:
the inference engine and the knowledge base.
• The knowledge base represents facts and rules.
• The inference engine applies the rules to the known
facts to deduce new facts or predictions.

Ch
The Architecture of Expert Systems

• Expert knowledge
derived from
human experts
• Purpose:
 Diagnose
illnesses
 Provide
recommendations
 Solve other
Problems

13
Examples of Expert System
based Rule Base systems

• For example, the GenAID system has about 10,000 rules


when it was first deployed.
• MYCIN was a joint venture between Dept. of Computer
Science and the Medical School of Stanford
University.
• MYCIN was designed to solve the problem of diagnosing
and recommending treatments for meningitis and blood
infections.
• Teaching Aids such as SCHOLAR which gives
Geography Tutorials and
• SOPHIE which teaches how to detect breakdown in
electrical circuits.
Ch
14
Case-based reasoning?

 An approach to building KBS which is radically


different to the rule-based and other knowledge-
representation approaches we have seen so
far.
 The principle is to find a solution which has been
shown to solve problems like your current
problem in the past, and adapt it so that it solves
the current problem.
Example of case based
reasoning?

• Oops the car stopped.


 What could have gone wrong?
• Aah.. Last time it happened, there was no petrol.
 Is there petrol?
 Yes.
 Oh but wait I remember the tyre was punctured (ban
bocor)
• This is the normal thought process of a human
when faced with a problem which is similar to a
problem he/she had faced before.
Ch
How do we solve problems?

• By knowing the steps to apply


 from symptoms/gejala to a plausible diagnosis
• But not always applying causal knowledge
 sebab - akibat
• How does an expert solve problems?
 uses same “book learning” as a novice
 but quickly selects the right knowledge to
apply
• Heuristic knowledge (“rules of thumb”)
 “I don’t know why this works but it does and so I’ll use it
again!” Ch
How a CBR system works:
the knowledgebase

• The knowledge base contains a collection of


representative cases (of faults, say, if the system
is concerned with fault diagnosis), with their
 symptoms,
 causes,
 and treatments.
How a CBR system works:
the process

• The user is instructed to provide the (relevant)


features of the current case.
• The similarity between this set of features, and the
features characteristic of each of the stored
cases is calculated, and the best match is
chosen.
Example -- Technical Diagnosis of
Car Faults

Ch
How a CBR system works:
the process

Current
Case
284742

134752 135753 134744 284743 144702


Case 1 Case 2 Case 3 Case 4 Case 5
How a CBR system works:
the process

Current
Case
284742

134752 135753 134744 284743 144702


Case 1 Case 2 Case 3 Case 4 Case 5
How a CBR system works:
the process

Current
Case
284742

134752 135753 134744 284743 144702


Case 1 Case 2 Case 3 Case 4 Case 5
How a CBR system works:
the process

Current
Case
284742

134752 135753 134744 284743 144702


Case 1 Case 2 Case 3 Case 4 Case 5
How a CBR system works:
the process

Current
Case
284742

134752 135753 134744 284743 144702


Case 1 Case 2 Case 3 Case 4 Case 5
How a CBR system works:
the process

Current
Case
284742

134752 135753 134744 284743 144702


Case 1 Case 2 Case 3 Case 4 Case 5
How a CBR system works:
the process

• The features which have been identified as


important in the stored cases, and which the user
is asked about, are known as “indices”.
• Each has a value. In the example I just showed
you, each was represented by a number.
How a CBR system works:
the process

• If necessary, this case is adapted so that it is a


better match for the current circumstances.
• The case is then presented as the solution, with
the opportunity to examine the 'precedent'
case.
How a CBR system works

• The sequence of operations,


for a simple CBR system:
1) assign indices
2) retrieve a similar case
Flow chart for a simple CBR
system

Input

ase memory 1. Assign indices


C Indexing rules

2. Retrieve Similarity
Case memory
metrics

Output
How a CBR system works

• The sequence of operations,


for a “full-blown” CBR system:
1) assign indices
2) retrieve a similar case
3) modify the past case
4) test the case
5a) assign indices to this new case,
and store as a working solution
OR
5b) explain failure, repair the solution,
and test again.
Flow chart for a full-blown CBR
system
Input
Indexing rules
1. Assign indices

2. Retrieve Similarity
Case memory
metrics

Modification
5b. Store 3. Modify rules

5a. Assign indices 4. Test


6b. Repair
Failed
Working solution
Repair rules
solution 6a. Explain
Available techniques for
case memory organisation

• Memory organisation by:


linear ("flat") case memory
 case hierarchy
nested cases
decision-tree orientated memory
 knowledge-guided indexing
Available techniques for
case retrieval

• Retrieval by:
Nearest neighbour case matching
Weighted nearest neighbour case matching
Decision tree methods
Knowledge-guided retrieval
• The last four memory organisation approaches,
and the last two retrieval approaches, might
be thought of as hybrid systems.
“Nearest neighbour”
algorithm: an example

• Suppose that we have a sick soyabean plant,


and we wish to discover which of a number of
known specimens of sick soyabean plants it is
most like.
Ch
Ch
Ch
“Nearest neighbour”
algorithm: an example

• Choose (let’s say) three characteristics of the


leaves that can be represented as numbers:
 Amount of the leaf that is covered by the
discolouration
 Lightness of the discoloured parts of the
leaf
 Lightness of the remaining parts of the
leaf.
“Nearest neighbour”
algorithm: an example

• Suppose that the first two cases to be matched


are:
• case 1: coverage - 8
lightness-1 - 4
lightness-2 - 6
• case 2: coverage - 10
lightness-1 - 7
lightness-1 - 6
“Nearest neighbour”
algorithm

• This can be treated as two points in three-


dimensional space:
• x, y, z coordinates of case 1: (8, 4, 6)
• x, y, z coordinates of case 2: (10, 7, 6)
“Nearest neighbour”
algorithm

10
9
8
7
6
5
4
3
2
1
1 x
3
2
0 1 2 3 4 5 6 7 8 9
4 10
7
6
5
A system of
3-dimensional
8
9
10

co-ordinates Ch
z 39
“Nearest neighbour”
algorithm

10
9
8
7
6
5
4
 - case 1
3
2
1
1  x
3
2
0 1 2 3 4 5 6 7 8 9
4 10
7
6
5
The 1st case
represented
8
9
10

as a point Ch
z 40
“Nearest neighbour”
algorithm

10
9
8
7
6
5
4
3 
2  - case 2
1
1 x
3
2
0 1 2 3 4 5 6 7 8 9
4 10
7
6
5
The 2nd case
10
9 8 represented
as a point Ch
z
“Nearest neighbour”
algorithm

10
9
8
7
6
5  - case 1
4
3   - case 2
2
1
1  x
3
2
0 1 2 3 4 5 6 7 8 9
4 10
7
6
5
The two cases
10
9 8 represented
as points Ch
z
“Nearest neighbour”
algorithm

10
9
8
7
6
5  - case 1
4
3   - case 2
2
1
1  x
3
2
0 1 2 3 4 5 6 7 8 9
4 10
7
6
5
The distance
between the
8
9
10

two cases Ch
z
“Nearest neighbour”
algorithm

10
9
8
 - case 1
7
6
 - case 2
5
4
 - case 3
3 
2
1
1  x
3
2
0 1 2 3 4 5 6 7 8 9
4 10
6 5
Adding a
7

third case:
8
9
10

(2, 3, 9) Ch
z
“Nearest neighbour”
algorithm

• There is a simple formula that tells you the


distance between two points in 3-dimensional
space.
• To find out whether case 1 is more similar to
case 2 or to case 3, you simply calculate the two
distances, and pick the smaller of the two.
“Nearest neighbour”
algorithm

• To find out which of a whole series of cases


case 1 is most similar to, calculate the distance
from case 1 to each of them, and pick the
smallest figure.
“Nearest neighbour”
algorithm

• Suppose it was 4 features, or 7, or 100? Would


you have to draw 4-dimensional or 7-
dimensional or 100-dimensional graphs?
• No, it’s simply necessary to have a formula for
calculating distances in 4, or 7, or 100-
dimensional space, and such formulae are
readily available.
Case adaptation

 "Fixing" inconsistencies between diagnosis and


symptoms.
 Techniques:
 the end user does it
 knowledge-based (qualitative reasoning, etc)
 a fixed procedure.
Case adaptation

• Note that there is a problem about updating the


case-base with adapted cases.
 Since the new case isn’t exactly like any of the cases in
the case-base, it can’t really be said to have been solved
by the expert judgement that was used to build the
case- base in the first place.
 There is a real chance that the conclusion that the system
came to is wrong in this case
 If wrongly concluded cases are added to the case-base, it
becomes progressively degraded.
Case adaptation

• Typically, the procedure is to put fresh cases


into a special file, and have the Domain Expert
pass judgement on them before they are added
to the case-base.
Appropriate domains

• CBR is suitable:
 when the domain is broad but shallow.
 when experience rather than theory is the
primary source of information.
 when the requirement is for the best available
solution, rather than a guaranteed exact
solution.
 when solutions are reusable, rather than unique
to each situation.
Example of a successful
system

• CBR is particularly used for help-


desk applications.
• For instance the COMPAQ SMART system.
Example of a successful
system

• The problem was that:


Thousands of customers were calling
Compaq directly every day, requesting
support.
Many of the staff were new; there was a
major training problem.
There was a need for consistent &
accurate answers and responses
There was a need for retention of
corporate knowledge.
Example of a successful
system

• The COMPAQ SMART system, once developed


and installed, succeeded in solving 85-95% of
calls.
• Typical time to solve a problem was less than 2
minutes.
Advantages of CBR

• Case-based reasoning:
tends to focus on the problem's
essential features.
can solve problems in domains that are
only partially understood.
can provide solutions when no
algorithmic method is available.
can interpret open-ended and ill-defined
concepts.
Steps in building a case-
based reasoning
system
1. Obtain data for cases.
2. Design cases based on data.
3. Determine the case memory structure.
4. Decide the case retrieval method.
5. Decide whether a case adaptation procedure
is appropriate (and, if so, implement it).
6. Develop the rest of the system (e.g. the
user interface).
Some currently-available
CBR tools (with
vendors)
 Esteem (Esteem Software)
 CBR Express & CBR v.2.0 (Inference)
 ReMind (Intelligent Applications, Cognitive
Systems)
 ReCall (ISoft)
 KATE-CBR (Acknosoft)
 Some of these are UK products, some
American,
some French.
Advantage of CBR over Rule
Base

• CBR is selected when the relationship between the case


attributes and the solution or outcome is not understood
well enough to represent in rules.
• CBR systems are advantageous when the ratio of cases
that are “exceptions to the rule” is high, as rule-based
systems become impractical in such applications.
• In such situation,
 CBR is convenient to incorporate the solution of a
newly entered case.
 CBR provides users with steps to combine and derive a
solution from the collection of retrieved solutions.
Ch
58
Decision support system

• A Decision Support System (DSS) is an interactive,


flexible, and adaptable a computer-based
information system, specially developed
for supporting the solution of a non-structured
management problem for improved decision making.
• A properly designed DSS help decision makers
compile useful information from a combination of
raw data, documents, personal knowledge, or
business models to identify and solve problems and
make decisions.
59
Capabilities of DSS

1. Provide support in semi-structured and unstructured


situations
2. Support for various managerial levels
3. Support to individuals and groups
4. Support to interdependent and/or sequential decisions
5. Support all phases of the decision-making process
6. Support a variety of decision-making processes and styles
7. Are adaptive
8. Have user friendly interfaces
9.

Ch
60
Capabilities ...

1. Goal is to improve the effectiveness of decision


making types
2. The decision maker controls the decision-making
process
3. End-users can build simple systems
4. Utilizes models for analysis
5. Provides access to a variety of data sources, formats,
and types
Decision makers can make better, more consistent
decisions in a timely manner
Decision Support Systems and Intelligent Systems, Efraim Turban and Jay E. Aronson
Ch
Copyright 1998, Prentice Hall, Upper Saddle River, NJ
Fault diagnosis Systems

• Fault diagnosis is increasingly becoming a major


emphasis for the development of knowledge
applications systems
• CABER at Lockheed Martin Corporation is one
of the earliest successful implementations of
knowledge application systems for the diagnosis
and recovery of faults in large multistation
milling machine tools

Ch
Fault diagnosis ...

• The milling machines are equipped with self-


diagnostic capabilities, typically they resolved only
20 percent to 40 percent of the systems faults.
• The field service engineers collected over
10,000 records for the creation of the case
library that supports CABER system,
• CABER augmented the self-diagnostic
capabilities of the milling machine, which
provided junior field- service engineers with the
necessary tools to resolve the fault and reduce
machine downtime.
Ch
Help Desk Systems

• A system that is developed to support customers as


help desk workers
• For example Compaq Computer Corporation
implemented a help desk support technology named
SMART to assist help desk employees track calls
and resolve customer service problems
• SMART is an integrated call-tracking and problem-
solving system, supported by hundreds of cases that
help resolve diagnostic problems resulting from the use
of Compaq products

Ch
Help Desk ....

• The system automatically retrieves from the


case library historical cases similar to the one
currently faced by the customer
• The customer service representative then uses
that solution to help customers solve the
problem at hand.
• SMART developers reported an increase from
50 percent to 87 percent of the problems that
could be resolved directly by the first level of
customer support.
Ch
Limitations of knowledge
application systems

• Typically developed to serve a task-specific domain


problem, and not integrated with the organization’s
enterprise systems.
• Security: cases may include sensitive information.
• Scalability: must represent a large enough
number of cases
• Speed: as the size of the case library grows to a more
comprehensive representation of real environments,
computing and searching costs will also increase.
• May not be able to solve all the problems that come
across,
in particular, increasingly complex environments
Ch

You might also like