Chapter 4 Knowledge Representation 323 670 Artificial Intelligence

Chapter 4 Knowledge Representation 323 -670 Artificial Intelligence ��. ����������

Representations and Mappings Ø Fact Ø Representations of facts Spot is a dog(spot) X: dog(X) hastail(X) All dogs have tails. Every dog has a tail. hastail(spot) Spot has a tail. 323 -670 Artificial Intelligence Lecture 13 Page 2

Mapping between Facts and Representations 323 -670 Artificial Intelligence Lecture 13 Page 3

Representation Of Facts generate appropriate final facts 323 -670 Artificial Intelligence Lecture 13 Page 4

Approaches to Knowledge Representation Ø Representational Adequacy : all kinds of knowledge in the domain Ø Inferential Adequacy : derive new structures corresponding to new knowledge inferred from old Ø Inferential Efficiency : incorporate into the knowledge structure for the most promising direction Ø Acquisitional Efficiency : acquire new knowledge easily 323 -670 Artificial Intelligence Lecture 13 Page 5

Simple Relational Knowledge ØRepresent set of attribute and associate value ØUsed in database system ØWeak inferential capacity ØUse as input for more powerful inference engines 323 -670 Artificial Intelligence Lecture 13 Page 6

Inheritable Knowledge ØProperty inheritance : element of specific classes inherit attributes and values from more general classes ØGeneralization hierarchy ØSlot-and-filler structure ØSemantic network : chapter 9 ØFrames : chapter 9 323 -670 Artificial Intelligence Lecture 13 Page 7

Inheritable Knowledge ØBaseball knowledge Øisa : show class inclusion Øinstance : show class membership 323 -670 Artificial Intelligence Lecture 13 Page 8