Structured Knowledge Chapter 7

2 Logic Notations Does logic represent well knowledge in structures?

3 Logic Notations Frege’s Begriffsschrift (concept writing) : assert P not P if P then Q for every x, P(x) P P Q P P(x) x

4 Logic Notations Frege’s Begriffsschrift (concept writing) : “Every ball is red” “Some ball is red” red(x) x ball(x) red(x) x ball(x)

5 Logic Notations Algebraic notation - Peirce, 1883: Universal quantifier:  x P x Existential quantifier:  x P x

6 Logic Notations Algebraic notation - Peirce, 1883: “Every ball is red”:  x (ball x —  red x ) “Some ball is red”:  x (ball x red x )

7 Logic Notations Peano’s and later notation: “Every ball is red”: (  x)(ball(x)  red(x)) “Some ball is red”: (  x)(ball(x)  red(x))

8 Logic Notations Existential graphs - Peirce, 1897: Existential quantifier: a link structure of bars, called line of identity, represents  Conjunction: the juxtaposition of two graphs represents  Negation: an oval enclosure represents ~

9 Logic Notations “If a farmer owns a donkey, then he beats it”: farmerownsdonkey beats

10 Logic Notations EG’s rules of inferences: Erasure: in a positive context, any graph may be erased. Insertion: in a negative context, any graph may be inserted. Iteration: a copy of a graph may be written in the same context or any nested context. Deiteration: any graph may be erased if a copy of its occurs in the same context or a containing context. Double negation: two negations with nothing between them may be erased or inserted.

11 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) is valid

12 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) is valid r p s q p qr s

13 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) Erasure: in a positive context, any graph may be erased. Insertion: in a negative context, any graph may be inserted. Iteration: a copy of a graph may be written in the same context or any nested context. Deiteration: any graph may be erased if a copy of its occurs in the same context or a containing context. Double negation: two negations with nothing between them may be erased or inserted.

14 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) Erasure: in a positive context, any graph may be erased. Insertion: in a negative context, any graph may be inserted. Iteration: a copy of a graph may be written in the same context or any nested context. Deiteration: any graph may be erased if a copy of its occurs in the same context or a containing context. Double negation: two negations with nothing between them may be erased or inserted. r p s q

15 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) Erasure: in a positive context, any graph may be erased. Insertion: in a negative context, any graph may be inserted. Iteration: a copy of a graph may be written in the same context or any nested context. Deiteration: any graph may be erased if a copy of its occurs in the same context or a containing context. Double negation: two negations with nothing between them may be erased or inserted. r p s q r p s q r p

16 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) Erasure: in a positive context, any graph may be erased. Insertion: in a negative context, any graph may be inserted. Iteration: a copy of a graph may be written in the same context or any nested context. Deiteration: any graph may be erased if a copy of its occurs in the same context or a containing context. Double negation: two negations with nothing between them may be erased or inserted. r p s q r p r p s q r p q

17 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) Erasure: in a positive context, any graph may be erased. Insertion: in a negative context, any graph may be inserted. Iteration: a copy of a graph may be written in the same context or any nested context. Deiteration: any graph may be erased if a copy of its occurs in the same context or a containing context. Double negation: two negations with nothing between them may be erased or inserted. r p s q p q s q r r p s q r

18 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) Erasure: in a positive context, any graph may be erased. Insertion: in a negative context, any graph may be inserted. Iteration: a copy of a graph may be written in the same context or any nested context. Deiteration: any graph may be erased if a copy of its occurs in the same context or a containing context. Double negation: two negations with nothing between them may be erased or inserted. r p s q p q s q r r p s q r s

19 Existential Graphs Prove: ((p  r)  (q  s))  ((p q)  (r s)) r p s q r p s q r p r p s q r p q r p s q s q r r p s q r s

20 Existential Graphs  -graphs: propositional logic  -graphs: first-order logic  -graphs: high-order and modal logic

21 Semantic Nets Since the late 1950s dozens of different versions of semantic networks have been proposed, with various terminologies and notations. The main ideas: For representing knowledge in structures The meaning of a concept comes from the ways it is connected to other concepts Labelled nodes representing concepts are connected by labelled arcs representing relations

22 Semantic Nets Mammal Person Owen Nose RedLiverpool isa instance has-part uniform color team person(Owen)  instance(Owen, Person) team(Owen, Liverpool)

23 Semantic Nets John H1H2 height Bill height greater-than 1.80 value

24 Semantic Nets Johng Give agent Book b instance object “John gives Mary a book” Mary beneficiary give(John, Mary, book)

25 Frames A vague paradigm - to organize knowledge in high- level structures “A Framework for Representing Knowledge” - Minsky, 1974 Knowledge is encoded in packets, called frames (single frames in a film) Frame name + slots

26 Frames Animals Alive Flies T F Birds Legs Flies 2 T Mammals Legs 4 Penguins Flies F CatsBats Legs Flies 2 T Opus Name Friend Opus Bill Name Friend Bill Pat Name Pat instance isa

27 Frames Hybrid systems: Frame component: to define terminologies (predicates and terms) Predicate calculus component: to describe individual objects and rules

28 Conceptual Graphs Sowa, J.F Conceptual Structures: Information Processing in Mind and Machine. CG = a combination of Perice’s EGs and semantic networks.

29 Conceptual Graphs 1968: term paper to Marvin Minsky at Harvard. 1970's: seriously working on CGs 1976: first paper on CGs : meeting with Norman Foo, finding Peirce’s EGs 1984: the book coming out CG homepage:

30 Simple Conceptual Graphs CAT: tunaMAT: *On 12 conceptrelationconcept type (class) relation type individual referentgeneric referent

31 Ontology Ontology: the study of "being" or existence An ontology = "A catalog of types of things that are assumed to exist in a domain of interest" (Sowa, 2000) An ontology = "The arrangement of kinds of things into types and categories with a well-defined structure" (Passin 2004)

32 Ontology top-level categories domain-specific

33 Ontology Being Substance Accident Property Relation Inherence Directedness MovementIntermediacyQuantityQuality Aristotle's categories Containment Activity Passivity HavingSituated SpatialTemporal

34 Ontology Geographical-Feature AreaLine On-LandOn-Water Road Border Mountain Terrain Block Geographical categories Railroad Country Wetland Point Power-Line River Heliport Town Dam Bridge Airstrip

35 Ontology Relation

36 Ontology Relation

37 Ontology ANIMAL FOOD Eat PERSON: johnCAKE: * Eat

38 CG Projection PERSON: johnPERSON: *Has-Relative 12 PERSON: johnWOMAN: maryHas-Wife 12

39 Nested Conceptual Graphs It is not true that cat Tuna is on a mat. CAT: tunaMAT: * On Neg CAT: tunaMAT: * On 

40 Nested Conceptual Graphs Every cat is on a mat. CAT: *MAT: * On  CAT: *  coreference link

41 Nested Conceptual Graphs Julian could not fly to Mars. PERSON: julianPLANET: mars Fly-To  Poss Past

42 Nested Conceptual Graphs Tom believes that Mary wants to marry a sailor. PERSON: julianPLANET: mars Fly-To  Poss Past

43 Exercises Reading: Sowa, J.F Knowledge Representation: Logical, Philosophical, and Computational Foundations (Section 1.1: history of logic). Way, E.C Conceptual Graphs – Past, Present, and Future. Procs. of ICCS'94.