Turban Chapter 12 Knowledge Representation KR

1
Turban Chapter 12 Knowledge Representation (KR)

• A good knowledge representation naturally
  represents the problem domain
• An unintelligible knowledge representation is
  wrong
• KR schemas
• Can be programmed
• Make knowledge usable for inference
• PRODUCTION RULES
• FRAMES
• OTHERS.

2
Other means of KR

• Propositional Logic
• A fact Witches ride brooms
• Another fact Hazel rides a broom
• A conclusion - Hazel is a witch
• Predicate Calculus
• Predicate(object, .)
• Is a (Hazel, Witch)
• Scripts
• Props, Roles, Track, Scenes

3
Other means of KR (2)

• Lists Outlines
• Decision Tables (Induction)
• Attributes (independent variables)
• Conclusions (dependent variables)
• Decision Trees
• Y/N T/F (Does Hazel ride a broom?)
• Object, Attribute, Value (OAV)
• Hazel, broom, yes

4
Other means of KR (3)

• Default Logic
• http//www.cs.engr.uky.edu/mirek/computing-with-d
  l.html
• Knowledge Map
• Conceptual not actual
• Semantic Networks
• Nodes
• Links
• Object, Class, inheritance

rides
Hazel
Broom
5
Production Rules

• If condition Then result
• If Hazel rides a broom then Hazel witch
• Knowledge
• If Hazel sweeps Then Hazel has a broom
• Inference
• If Hazel has a broom Then Determine if she rides
  it

6
Production Rules (2)

• Each production rule in a knowledge base
  represents an autonomous chunk of expertise
• When combined and fed to the inference engine,
  the set of rules behaves synergistically
• Rules can be viewed as a simulation of the
  cognitive behavior of human experts
• Rules represent a model of actual human behavior

7
Frames

• Data Structure
• Objects/Entities
• Structure
• Slots facts or procedures
• Facets
• Values, range, defaults
• Demons - If added, if needed.

8
Frames (2)
Name Nose Is A
Hazel Wart
Sex Politics Religion
Female Democrat Witch
Group Vehicle Nose
Witches Broom Wart
(instance)
9
Frames (3)

• Data structure that includes all the knowledge
  about a particular object
• Knowledge organized in a hierarchy for diagnosis
  of knowledge independence
• Form of object-oriented programming for AI and
  ES.
• Each Frame Describes One Object

10
Experimental KR

• CYC www.cyc.com

11
Uncertainty

• When a user cannot provide a definite answer
• Imprecise knowledge
• Incomplete information
• Techniques
• Bayesian Statistics
• Dempster and Shafer's Belief Functions
• Fuzzy Sets