Kuliah Sistem Pakar
Kuliah Sistem Pakar
Pertemuan V
Pertemuan V
“Representasi Pengetahuan”
Tujuan Pembelajaran
Tujuan Pembelajaran
Mengerti perang proses RPL terhadap Rekayasa Pengetahuan Mengerti Representasi Pengetahuan, tipe-tupe
Mengetahui Tipe – Tipe Representasi Pengetahuan
Proses Rekayasa Pengetahuan
Proses Rekayasa Pengetahuan
(
(
Knowledge Engineering Process)
Knowledge Engineering Process)
ValidasiPengetahuan PengetahuanSumber
Representasi Pengetahuan Basis
Pengetahuan
Justifkasi Penjelasan
Inferensi
Akuisisi Pengetahuan
Knowledge Representation
Knowledge Representation
Knowledge Representation
Knowledge Representation
is concerned with
is concerned with
storing large bodies of useful information in a
storing large bodies of useful information in a
symbolic format.
symbolic format.
Most commercial ES are
Most commercial ES are
rule-based systems
rule-based systems
where the information is stored as rules.
where the information is stored as rules.
Frames may also be used to complement rule-based
Frames may also be used to complement rule-based
systems.
Tipe-tipe Pengetahuan berdasar
Tipe-tipe Pengetahuan berdasar
Sumber
Sumber
Deep Knowledge
Deep Knowledge
(formal knowledge)
(formal knowledge)
Shallow /Surface Knowledge
Shallow /Surface Knowledge
(non formal knowledge)
Penjelasan ………
Penjelasan ………
Deep Deep knowledge knowledge atauatau pengetahuan pengetahuan formal,formal, pengetahuan bersifat umum yang
pengetahuan bersifat umum yang terdapat dalam sumber terdapat dalam sumber pengetahuan tertentu (buku, jurnal, buletin ilmiah dsb)
pengetahuan tertentu (buku, jurnal, buletin ilmiah dsb)
dan dapat diterapkan dalam tugas maupun kondisi
dan dapat diterapkan dalam tugas maupun kondisi
berbeda.
berbeda.
Shallow knowledge Shallow knowledge atauatau pengetahuan non formal, pengetahuan non formal, pengetahuan-pengetahuan praktis dalam bidang tertentu
pengetahuan-pengetahuan praktis dalam bidang tertentu
yang diperoleh seorang pakar pengalamannya pada
yang diperoleh seorang pakar pengalamannya pada
bidang dalam jangka waktu cukup lama.
Pengetahuan HeuristikPengetahuan Heuristik
Pengetahuan ProseduralPengetahuan Prosedural
Pengetahuan DeklaratifPengetahuan Deklaratif
Tipe-tipe Pengetahuan berdasar Cara
Tipe-tipe Pengetahuan berdasar Cara
Representasi Pengetahuan
Representasi Pengetahuan
Propotional LogicPropotional Logic (logika proposional)(logika proposional) Semantic NetworkSemantic Network (jaringan semantik)(jaringan semantik) Script, List, Table, dan TreeScript, List, Table, dan Tree
Object, Attribute, dan ValuesObject, Attribute, dan Values
Representation in Logic and
Representation in Logic and
Other Schemas
Other Schemas
General form of any logical process
General form of any logical process
Inputs (Premises)
Inputs (Premises)
Premises used by the logical process to
Premises used by the logical process to
create the output, consisting of
create the output, consisting of
conclusions (inferences)
conclusions (inferences)
Facts known true can be used to derive
Facts known true can be used to derive
new facts that also must be true
Two Basic Forms of Computational Logic
Two Basic Forms of Computational Logic
Symbols represent propositions, premises or
Symbols represent propositions, premises or
conclusions
conclusions
Statement: A = The mail carrier comes Monday
Statement: A = The mail carrier comes Monday
through Friday.
through Friday.
Statement: B = Today is Sunday.
Statement: B = Today is Sunday.
Conclusion: C = The mail carrier will not come
Conclusion: C = The mail carrier will not come
today.
today.
Propositional logic: limited in representing
Propositional logic: limited in representing
real-world knowledge
Propositional Logic
Propositional Logic
A proposition is a statement that is either A proposition is a statement that is either truetrue or or
false
false
Once known, it becomes a premise that can be used Once known, it becomes a premise that can be used
to derive new propositions or inferences
to derive new propositions or inferences
Rules are used to determine the truth (T) or falsity Rules are used to determine the truth (T) or falsity
(F) of the new proposition
Propotional Logic
Propotional Logic
Logic dapat digunakan untuk melakukan penalaran :
Logic dapat digunakan untuk melakukan penalaran :
Contoh :
Pernyataan B = Hari ini hari Minggu
Pernyataan B = Hari ini hari Minggu
Kesimpulan C = Pak Pos tidak akan datang hari ini
Kesimpulan C = Pak Pos tidak akan datang hari ini
Predicate Calculus
Predicate Calculus
Predicate logic breaks a statement down into Predicate logic breaks a statement down into
component parts, an object, object characteristic or
component parts, an object, object characteristic or
some object assertion
some object assertion
Predicate calculus uses variables and functions of Predicate calculus uses variables and functions of variables in a symbolic logic statement
variables in a symbolic logic statement
Predicate calculus is the basis for Prolog Predicate calculus is the basis for Prolog (PROgramming in LOGic)
(PROgramming in LOGic)
Prolog Statement ExamplesProlog Statement Examples
comes_on(mail_carrier, monday).comes_on(mail_carrier, monday). likes(jay, chocolate).likes(jay, chocolate).
(Note - the period “.” is part of the statement)
Merupakan gambaran pengetahuan
Merupakan gambaran pengetahuan
berbentuk grafs dan menunjukkan
berbentuk grafs dan menunjukkan
hubungan antar berbagai obyek.
hubungan antar berbagai obyek.
Obyek, berupa benda
Obyek, berupa benda
atau
atau
peristiwa
peristiwa
Nodes Obyek
Nodes Obyek
Arc (Link) Keterhubungan
Arc (Link) Keterhubungan
(Relationships)
(Relationships)
*
* is a
is a
* has a
* has a
Scripts
Scripts
SCRIPT
SCRIPT
,
,
skema skema representasi representasi pengetahuan pengetahuan yang yang menggambarkan urutan dari kejadian. Elemen-elemenmenggambarkan urutan dari kejadian. Elemen-elemen
script terdiri dari :
script terdiri dari :
Elements include Elements include
Entry ConditionsEntry Conditions PropsProps
RolesRoles Tracks Tracks ScenesScenes
LIST,LIST,
daftar tertulis dari item-item yang saling
daftar tertulis dari item-item yang saling
berhubungan.
berhubungan.
Umumnya digunakan untuk merepresentasikan Umumnya digunakan untuk merepresentasikan
hirarki pengetahuan dimana suatu obyek
hirarki pengetahuan dimana suatu obyek
dikelompokan, dikategorikan sesuai dengan
dikelompokan, dikategorikan sesuai dengan
Rank or Rank or
produk-produk dalam suatu katalog.
produk-produk dalam suatu katalog.
List
DECISION TABLE,DECISION TABLE, pengetahuan yang diatur dalam pengetahuan yang diatur dalam
format lembar kerja atau
format lembar kerja atau spreadsheetspreadsheet, menggunakan , menggunakan kolom dan baris.
kolom dan baris.
Attribute List
Attribute List
Conclusion List
Conclusion List
Different attribute configurations are matched against
Different attribute configurations are matched against
the conclusion
the conclusion
Contoh :… ?Contoh :… ?
Decision Tabel
Decision Trees
Decision Trees
DECISION TREEDECISION TREE,, treetree yang berhubungan dengan yang berhubungan dengan decision decision table
table namun sering digunakan dalam analisis sistem komputer namun sering digunakan dalam analisis sistem komputer (bukan sistem AI).
(bukan sistem AI).
Contoh :… ?Contoh :… ?
Related to tables Related to tables
Similar to decision trees in decision theorySimilar to decision trees in decision theory
Can simplify the knowledge acquisition processCan simplify the knowledge acquisition process Knowledge diagramming is frequently more Knowledge diagramming is frequently more
natural to experts than formal representation
natural to experts than formal representation
methods
Object, Attribute, Values
Object, Attribute, Values
OBJECT
OBJECT : :
OBJECTOBJECT dapat berupa fisik atau konsepsi. dapat berupa fisik atau konsepsi.
ATTRIBUTE
ATTRIBUTE : :
ATTRIBUTEATTRIBUTE adalah karakteristik dari object. adalah karakteristik dari object.
VALUES
VALUES : :
VALUESVALUES adalah ukuran spesifik dari attribute dalam adalah ukuran spesifik dari attribute dalam
situasi tertentu
Object Attribute Values
Object Attribute Values
Rumah
Rumah Kamar tidurKamar tidur 2,3,4, dsb.2,3,4, dsb.
Rumah
Rumah WarnaWarna Hijau, Putih, Hijau, Putih, Coklat dsb.
Coklat dsb.
Diterima di
Diterima di
Universitas
Universitas Nilai Ujian masuk
Nilai Ujian masuk A, B, C atau DA, B, C atau D
Pengendalian
Pengendalian
persedian
persedian Level persediaan
Level persediaan 15, 20, 25, 35, 15, 20, 25, 35, dsb.
dsb.
Kamar tidur
Kamar tidur UkuranUkuran 3x4, 5x6, 4x5, 3x4, 5x6, 4x5, dsb.
Production Rules
Production Rules
PRODUCTION RULES:
PRODUCTION RULES:
Production system dikembangkan oleh
Production system dikembangkan oleh
Newell dan Simon sebagai model dari
Newell dan Simon sebagai model dari
kognisi manusia. Ide dasar dari sistem ini
kognisi manusia. Ide dasar dari sistem ini
adalah pengetahuan digambarkan sebagai
adalah pengetahuan digambarkan sebagai
production rules dalam bentuk
production rules dalam bentuk
pasangan
pasangan
kondisi-aksi
Production Rules
Production Rules
Condition-Action PairsCondition-Action Pairs
IF this condition (or premise or antecedent) IF this condition (or premise or antecedent)
occurs,
occurs,
THEN some action (or result, or conclusion, or THEN some action (or result, or conclusion, or
consequence) will (or should) occur
consequence) will (or should) occur
IF the stop light is red AND you have stopped, IF the stop light is red AND you have stopped,
THEN a right turn is OK
Each production rule in a knowledge base represents Each production rule in a knowledge base represents
an
an autonomous chunkautonomous chunk of expertise of expertise
When combined and fed to the inference engine, the When combined and fed to the inference engine, the
set of rules behaves synergistically
set of rules behaves synergistically
Rules can be viewed as a simulation of the cognitive Rules can be viewed as a simulation of the cognitive
behavior of human experts
behavior of human experts
Contoh : Production Rules
Contoh : Production Rules
RULE 1 :
RULE 1 :
JIKA konfik internasional mulai
JIKA konfik internasional mulai
MAKA harga emas naik
MAKA harga emas naik
RULE 2 :
RULE 2 :
JIKA laju infasi berkurang
JIKA laju infasi berkurang
MAKA harga emas turun
MAKA harga emas turun
RULE 3
RULE 3
:
:
JIKA konfik internasional
JIKA konfik internasional
berlangsung lebih dari tujuh
berlangsung lebih dari tujuh
hari
hari
dan
dan
JIKA konfik terjadi di Timur
JIKA konfik terjadi di Timur
Tengah
Tengah
MAKA beli emas
Production Rules
Production Rules
Condition-Action Pairs
Condition-Action Pairs
IF this condition (or premise or
IF this condition (or premise or
antecedent) occurs,
antecedent) occurs,
THEN some action (or result, or
THEN some action (or result, or
conclusion, or consequence) will (or
conclusion, or consequence) will (or
should) occur
should) occur
IF the stop light is red AND you have
IF the stop light is red AND you have
stopped, THEN a right turn is OK
Each production rule in a
Each production rule in a
knowledge base represents an
knowledge base represents an
autonomous chunk
autonomous chunk
of expertise
of expertise
When combined and fed to the
When combined and fed to the
inference engine, the set of rules
inference engine, the set of rules
behaves synergistically
behaves synergistically
Rules can be viewed as a
Rules can be viewed as a
simulation of the cognitive
simulation of the cognitive
behavior of human experts
behavior of human experts
Rules represent a
Rules represent a
model
model
of actual
of actual
human behavior
Forms of Rules
Forms of Rules
IF premise, THEN conclusionIF premise, THEN conclusion
IF your income is high, IF your income is high,
THEN your chance of being audited by the THEN your chance of being audited by the
IRS is high
IRS is high
Conclusion, IF premiseConclusion, IF premise
Your chance of being audited is high, IF Your chance of being audited is high, IF
your income is high
Inclusion of ELSEInclusion of ELSE
IF your income is high, OR your deductions are IF your income is high, OR your deductions are
unusual, THEN your chance of being audited by
unusual, THEN your chance of being audited by
the IRS is high, OR ELSE your chance of being
the IRS is high, OR ELSE your chance of being
audited is low
audited is low
More Complex RulesMore Complex Rules
IF credit rating is high AND salary is more than IF credit rating is high AND salary is more than
$30,000, OR assets are more than $75,000, AND
$30,000, OR assets are more than $75,000, AND
pay history is not "poor," THEN approve a loan up
pay history is not "poor," THEN approve a loan up
to $10,000, and list the loan in category "B.”
to $10,000, and list the loan in category "B.”
Action part may have more information: THEN Action part may have more information: THEN
"approve the loan" and "refer to an agent"
Frame
Frame
FRAMEFRAME adalah struktur data yang berisi semua adalah struktur data yang berisi semua
pengetahuan tentang obyek tertentu. Pengetahuan
pengetahuan tentang obyek tertentu. Pengetahuan
ini diatur dalam suatu struktur hirarkis khusus yang
ini diatur dalam suatu struktur hirarkis khusus yang
memperbolehkan diagnosis terhadap independensi
memperbolehkan diagnosis terhadap independensi
pengetahuan. Frame pada dasarnya adalah aplikasi
pengetahuan. Frame pada dasarnya adalah aplikasi
dari pemrograman berorientasi objek untuk AI dan
dari pemrograman berorientasi objek untuk AI dan
ES.
ES.
Setiap frame mendefinisikan satu objek, dan terdiri Setiap frame mendefinisikan satu objek, dan terdiri
dari dua elemen :
dari dua elemen : slotslot (menggambarkan rincian dan (menggambarkan rincian dan karakteristik obyek) dan
Frames
Frames
FrameFrame: Data structure that includes all the : Data structure that includes all the
knowledge about a particular object
knowledge about a particular object
Knowledge organized in a hierarchy for diagnosis of Knowledge organized in a hierarchy for diagnosis of
knowledge independence
knowledge independence
Form of Form of object-oriented programmingobject-oriented programming for AI and ES. for AI and ES. Each Frame Describes One ObjectEach Frame Describes One Object
Contoh Frame
Automobile Frame
Automobile Frame
Class of : Transportation
Class of : Transportation
Name of Manufacturer : Audi
Name of Manufacturer : Audi
Origin of Manufacturer : Germany
Origin of Manufacturer : Germany
Model : 5000 turbo
Wheelbase : 105.8 inches
Wheelbase : 105.8 inches
Number of doors : 4 (default)
Number of doors : 4 (default)
Transmission : 3-speed (automatic)
Transmission : 3-speed (automatic)
Number of wheels : 4 (default)
Number of wheels : 4 (default)
Gas mileage : 22 mpg average (procedural attachment)
Gas mileage : 22 mpg average (procedural attachment)
Engine Frame
Engine Frame
Cylinder bore : 3.19 inches
Cylinder bore : 3.19 inches
Cylinder stroke : 3.4 inches
Cylinder stroke : 3.4 inches
Compression ratio : 7.8 to 1
Compression ratio : 7.8 to 1
Fuel system : Injection with turbocharger
Fuel system : Injection with turbocharger
Horsepower : 140 hp
Horsepower : 140 hp
Torque : 160 ft/Lbs
Hirarki Frame (exp : Vehicle)
Hirarki Frame (exp : Vehicle)
Advantages and Disadvantages of Different Knowledge Representations
Scheme Advantages Disadvantages
Production
rules Simple syntax, easy to understand, simple
interpreter, highly modular, flexible (easy to add to or modify)
Hard to follow hierarchies, inefficient for large systems, not all knowledge can be expressed as rules, poor at representing structured descriptive knowledge
Semantic
networks Easy to follow hierarchy, easy to trace associations, flexible
Meaning attached to nodes might be ambiguous,
exception handling is difficult, difficult to program
Frames Expressive power, easy to set up slots for new properties and relations, easy to create specialized procedures, easy to include default information and detect missing values
Difficult to program,
difficult for inference, lack of inexpensive software
Formal logic Facts asserted independently of use, assurance that all and only valid consequences are asserted (precision),
completeness
Separation of
representation and