Artificial Intelligence

I: knowledge repre-
sentation
Outline
 Ontological engineering
 Categories and objects
 Actions, situations and events
 Mental events and mental objects
 The internet shopping world
 Reasoning systems for categories
 Reasoning with default information
 Truth maintenance systems

2
Ontological engineering
 How to create more general and flexible representations.
 Concepts like actions, time, physical object and beliefs
 Operates on a bigger scale than K.E.

 Define general framework of concepts

 Upper ontology
 Limitations of logic representation
 Red, green and yellow tomatoes: exceptions and
uncertainty

3
The upper ontology of the world

4
Difference with special-purpose
ontologies
 A general-purpose ontology should be applicable in more
or less any special-purpose domain.
 Add domain-specific axioms
 In any sufficiently demanding domain different areas of
knowledge need to be unified.
 Reasoning and problem solving could involve several
areas simultaneously
 What do we need to express?
Categories, Measures, Composite objects, Time, Space,
Change, Events, Processes, Physical Objects,
Substances, Mental Objects, Beliefs

5
Categories and objects
 KR requires the organisation of objects into categories
 Interaction at the level of the object
 Reasoning at the level of categories

 Categories play a role in predictions about objects

 Based on perceived properties
 Categories can be represented in two ways by FOL
 Predicates: apple(x)
 Reification of categories into objects: apples

 Category = set of its members

6
Category organization
 Relation = inheritance:
 All instance of food are edible, fruit is a subclass of
food and apples is a subclass of fruit then an apple
is edible.
 Defines a taxonomy

7
FOL and categories
 An object is a member of a category
 MemberOf(BB12,Basketballs)
 A category is a subclass of another category
 SubsetOf(Basketballs,Balls)
 All members of a category have some properties
  x (MemberOf(x,Basketballs)  Round(x))
 All members of a category can be recognized by some properties
  x (Orange(x)  Round(x)  Diameter(x)=9.5in 
MemberOf(x,Balls)  MemberOf(x,BasketBalls))
 A category as a whole has some properties
 MemberOf(Dogs,DomesticatedSpecies)

8
Relations between categories
 Two or more categories are disjoint if they have no members in
common:
 Disjoint(s)( c1,c2 c1  s  c2  s  c1  c2  Intersection(c1,c2) ={})
 Example; Disjoint({animals, vegetables})

 A set of categories s constitutes an exhaustive decomposition of a

category c if all members of the set c are covered by categories in s:
 E.D.(s,c)  ( i i  c   c2 c2  s  i  c2)
 Example: ExhaustiveDecomposition({Americans,
Canadian, Mexicans},NorthAmericans).

9
Relations between categories
 A partition is a disjoint exhaustive decomposition:
 Partition(s,c)  Disjoint(s)  E.D.(s,c)
 Example: Partition({Males,Females},Persons).

 Is ({Americans,Canadian, Mexicans},NorthAmericans) a
partition?
 Categories can be defined by providing necessary and
sufficient conditions for membership
  x Bachelor(x)  Male(x)  Adult(x)  Unmarried(x)

10
Natural kinds
 Many categories have no clear-cut definitions (chair, bush,
book).
 Tomatoes: sometimes green, red, yellow, black. Mostly
round.
 One solution: category Typical(Tomatoes).
  x, x  Typical(Tomatoes)  Red(x)  Spherical(x).
 We can write down useful facts about categories without
providing exact definitions.
 What about “bachelor”? Quine challenged the utility of the
notion of strict definition. We might question a statement
such as “the Pope is a bachelor”.

11
Physical composition
 One object may be part of another:
 PartOf(Bucharest,Romania)
 PartOf(Romania,EasternEurope)
 PartOf(EasternEurope,Europe)
 The PartOf predicate is transitive (and irreflexive), so we can infer that
PartOf(Bucharest,Europe)
 More generally:
  x PartOf(x,x)
  x,y,z PartOf(x,y)  PartOf(y,z)  PartOf(x,z)
 Often characterized by structural relations among parts.
 E.g. Biped(a) 

12
Measurements
 Objects have height, mass, cost, ....
Values that we assign to these are measures
 Combine Unit functions with a number: Length(L 1) =
Inches(1.5) = Centimeters(3.81).
 Conversion between units:
 i Centimeters(2.54 x i)=Inches(i).
 Some measures have no scale: Beauty, Difficulty, etc.
 Most important aspect of measures: is that they are
orderable.
 Don't care about the actual numbers. (An apple can have
deliciousness .9 or .1.)

13
Actions, events and situations
• Reasoning about outcome of
actions is central to KB-agent.
• How can we keep track of
location in FOL?
• Remember the multiple
copies in PL.
• Representing time by situations
(states resulting from the
execution of actions).
• Situation calculus

14
Actions, events and situations
 Situation calculus:
 Actions are logical terms
 Situations are logical terms
consiting of
 The initial situation I
 All situations resulting from
the action on I (=Result(a,s))
 Fluent are functions and
predicates that vary from
one situation to the next.
 E.g. Holding(G1, S0)
 Eternal predicates are also
allowed
 E.g. Gold(G1)

15
Actions, events and situations
 Results of action sequences
are determined by the
individual actions.
 Projection task: an SC agent
should be able to deduce the
outcome of a sequence of
actions.
 Planning task: find a
sequence that achieves a
desirable effect

16
Actions, events and situations

17
Describing change
 Simples Situation calculus requires two axioms to
describe change:
 Possibility axiom: when is it possible to do the action
At(Agent,x,s)  Adjacent(x,y)  Poss(Go(x,y),s)
 Effect axiom: describe changes due to action
Poss(Go(x,y),s)  At(Agent,y,Result(Go(x,y),s))
 What stays the same?
 Frame problem: how to represent all things that stay the
same?
 Frame axiom: describe non-changes due to actions

At(o,x,s)  (o  Agent)  Holding(o,s)  At(o,x,Result(Go(y,z),s))

18
Representational frame
problem
 If there are F fluents and A actions then we need AF
frame axioms to describe other objects are stationary
unless they are held.
 We write down the effect of each actions
 Solution; describe how each fluent changes over time
 Successor-state axiom:
Pos(a,s)  (At(Agent,y,Result(a,s))  (a = Go(x,y)) 
(At(Agent,y,s)  a  Go(y,z))
 Note that next state is completely specified by current
state.
 Each action effect is mentioned only once.

19
Other problems
 How to deal with secondary (implicit) effects?
 If the agent is carrying the gold and the agent moves then
the gold moves too.
 Ramification problem

 How to decide EFFICIENTLY whether fluents hold in

the future?
 Inferential frame problem.
 Extensions:
 Event calculus (when actions have a duration)
 Process categories

20
Mental events and objects
 So far, KB agents can have beliefs and deduce new beliefs
 What about knowledge about beliefs? What about
knowledge about the inference proces?
 Requires a model of the mental objects in someone’s head
and the processes that manipulate these objects.
 Relationships between agents and mental objects: believes,
knows, wants, …
 Believes(Lois,Flies(Superman)) with Flies(Superman) being
a function … a candidate for a mental object (reification).
 Agent can now reason about the beliefs of agents.

21
The internet shopping world
 A Knowledge Engineering example
 An agent that helps a buyer to find product offers on the
internet.
 IN = product description (precise or precise)
 OUT = list of webpages that offer the product for sale.

 Environment = WWW
 Percepts = web pages (character strings)
 Extracting useful information required.

22
The internet shopping world
 Find relevant product offers
RelevantOffer(page,url,query)  Relevant(page, url, query)  Offer(page)
 Write axioms to define Offer(x)
 Find relevant pages: Relevant(x,y,z) ?
 Start from an initial set of stores.
 What is a relevant category?
 What are relevant connected pages?
 Require rich category vocabulary.
 Synonymy and ambiguity
 How to retrieve pages: GetPage(url)?
 Procedural attachment
 Compare offers (information extraction).

23
Reasoning systems for categories
 How to organise and reason with categories?
 Semantic networks
 Visualize knowledge-base
 Efficient algorithms for category membership inference
 Description logics
 Formal language for constructing and combining
category definitions
 Efficient algorithms to decide subset and superset
relationships between categories.

24
Semantic Networks
 Logic vs. semantic networks
 Many variations
 All represent individual objects, categories of objects and
relationships among objects.
 Allows for inheritance reasoning
 Female persons inherit all properties from person.
 Cfr. OO programming.

 Inference of inverse links

 SisterOf vs. HasSister

25
Semantic network example

26
Semantic networks
 Drawbacks
 Links can only assert binary relations
 Can be resolved by reification of the proposition as
an event
 Representation of default values
 Enforced by the inheritance mechanism.

27
Description logics
 Are designed to describe defintions and properties about
categories
 A formalization of semantic networks
 Principal inference task is
 Subsumption: checking if one category is the subset of
another by comparing their definitions
 Classification: checking whether an object belongs to a
category.
 Consistency: whether the category membership criteria
are logically satisfiable.

28
Reasoning with Default Information
 “The following courses are offered: CS101, CS102,
CS106, EE101”
 Four (db)
 Assume that this information is complete (not asserted
ground atomic sentences are false)
= CLOSED WORLD ASSUMPTION
 Assume that distinct names refer to distinct objects

= UNIQUE NAMES ASSUMPTION

 Between one and infinity (logic)
 Does not make these assumptions
 Requires completion.

29
Truth maintenance systems
 Many of the inferences have default status rather than
being absolutely certain
 Inferred facts can be wrong and need to be retracted =
BELIEF REVISION.
 Assume KB contains sentence P and we want to execute
TELL(KB, P)
 To avoid contradiction: RETRACT(KB,P)
 But what about sentences inferred from P?
 Truth maintenance systems are designed to handle these
complications.

30