Knowledge Representation

1
Outline

 General ontology
 Categories and objects
 Events and processes
 Reasoning systems
 Internet shopping world
 Summary

2
Ontologies

 An ontology is a “vocabulary” and a “theory” of a

certain “part of reality”
 Special-purpose ontologies apply to restricted
domains (e.g. electronic circuits)
 General-purpose ontologies have wider
applicability across domains, i.e.
 Must include concepts that cover many
subdomains
 Cannot use special “short-cuts” (such
as ignoring time)
 Must allow unification of different types
of knowledge
 GP ontologies are useful in widening 3
Ontological engineering

 Representing a general-purpose ontology is a

difficult task called ontology engineering
 Existing GP ontologies have been created in
different ways:
 By team of trained ontologists
 By importing concepts from database(s)
 By extracting information from text documents
 By inviting anybody to enter commonsense
knowledge
 Ontological engineering has only been partially
successful, and few large AI systems are based on
GP ontologies (use special purpose ontologies)
4
Elements of a general ontology

 Categories of objects
 Measures of quantities
 Composite objects
 Time, space, and change
 Events and processes
 Physical objects
 Substances
 Mental objects and beliefs

5
 Top-level ontology of the world

Anything

AbstractObjects Events

Sets RepresentationObjects Intervals Places PhysicalObjects Processes

Numbers

Categories Sentences Measurements Things Stuff

Moments

Time Weight Animals Agents Solids Liquid Gas

s s

Humans

6
Upper Ontology

• The general framework of concepts is called an upper ontology

because of the convention of drawing graphs with the general
concepts at the top and the more specific concepts below them
• Of what use is an upper ontology?
• Consider the ontology for circuits that we studied
• It makes many simplifying assumptions: time is omitted completely; signals
are fixed and do not propagate; the structure of the circuit remains constant.
• more general ontology would consider signals at particular times, and would
include the wire lengths and propagation delays.
• This would allow us to simulate the timing properties of the circuit, and
indeed such simulations are often carried out by circuit designers.

7
Categories and objects
 Categories are used to classify objects
according to common properties or definitions
x x  Tomates  Re d (x )  Round (x )
 Categories can be represented by
 Predicates: Tomato(x)
 Objects: The constant Tomatoes represents set
of
tomatoes (reification)
 Roles of category
 Instance relations representations
(is- x 1  Tomatoes
a):
 Taxonomical hierarchies (Subset): T om a toes  F ru it

 Inheritance of properties
 (Exhaustive) decompositions
8
Categories using FOL

9
Properties of categories

• We say that two or more categories are disjoint if they have no members
in common.
• exhaustive decomposition
• A disjoint exhaustive decomposition is known as a partition.
• The following examples illustrate these three concepts:

• The predicates us to define these concepts are

• For example, a bachelor is an unmarried adult male:

10
Objects and substance

 Need to distinguish between substance and

discrete objects
 Substance (“stuff”)
 Mass nouns - not countable
 Intrinsic properties
 Part of a substance is (still) the same
substance
 Discrete objects (“things”)
 Count nouns - countable
 Extrinsic properties
 Parts are (generally) not of same category

11
Composite objects

 A composite object is an object that has other

objects as parts
 The PartOf relation defines the object
containment, and is transitive and
reflexive PartOf (x, y)  PartOf ( y, z)  PartOf (x, z)
PartOf (x, x)
 Objects can be grouped in PartOf
hierarchies,
similar to Subset hierarchies
 The structure of the composite object
describes how the parts are related
12
Composite objects
• For example, a biped has two legs attached to a body:

• For example, we might want to say “The apples in this bag weigh two
pounds.”
• we need a new concept, which we will call a bunch.
• For example, if the apples are Apple1, Apple2, and Apple3, then
BunchOf ({Apple1,Apple2,Apple3})
∀x x∈ s ⇒ PartOf (x, BunchOf (s))
∀ y [∀x x∈ s ⇒ PartOf (x, y)] ⇒ PartOf (BunchOf (s), y)
• logical minimization, which means defining an object as the smallest one
satisfying certain conditions.
13
Measurements
 Need to be able to represent properties
like height, mass, cost, etc. Values for such
properties are measures
 Unit functions represent and convert
measures
Length( L1)  Inches(1.5)  Centimeters(3.81)
l Centimeters(2.54  l)  Inches(l)
 Measures can be used to describe objects
Mass(Tomato1)  Ki log rams(0.16)
d d  Days  Duration(d )  Hours(24)
 Non-numerical measures can also be represen-
ted, but normally there is an order (e.g. >).
Used in qualitative physics 14
Measurements

• Comparative difficulty

e1 ∈ Exercises ∧ e2 ∈Exercises ∧ Wrote(Norvig, e1) ∧ Wrote(Russell,

e2) ⇒ Difficulty(e1) > Difficulty(e2)

e1 ∈ Exercises ∧ e2 ∈Exercises ∧ Difficulty(e1) > Difficulty(e2) ⇒

ExpectedScore(e1) < ExpectedScore(e2)

15
Objects: Things and stuff

• The real world can be seen as consisting of primitive objects (e.g.,

atomic particles) and composite objects built from them.
• There is, however, a significant portion of reality that seems to defy
any obvious individuation—division into distinct objects. We give this
portion the generic name stuff.
• count nouns, such as aardvarks, holes, and theorems, and mass
nouns, such as butter, water, and energy.
• To represent stuff properly, we begin with the obvious. We need to
have as objects in our ontology at least the gross “lumps” of stuff we
interact with.
b∈ Butter ∧ PartOf (p, b) ⇒ p ∈Butter
b∈ Butter ⇒ MeltingPoint(b,Centigrade(30))
• Intrinsic properties and extrinsic properties
16
Event calculus

 Event calculus: How to deal with change based on

representing points of time
 Reifies fluents and events
 A fluent: At(Bilal, Berkeley)
 The fluent is true at time t: T(At(Bilal, IQRA),t)
 Events are instances of event categories
E1  Flyings  Flyer(E1, Bilal)  Origin(E1, SF )  Destination(E1,
KHI)
 Event E1 took place over interval i
 Happens(E1, i)
 Time intervals represented by (start, end) pairs
 i = (t1, t2) 17
 Event calculus predicates

 T(f, t) Fluent f is true at time t

 Happens(e, i) Event e happens over interval i
 Initiates(e, f, t) Event e causes fluent f to start at t
 Terminates(e, f, t) Event e causes f to cease at t
 Clipped(f, t) Fluent f ceases to be true in int. i
 Restored(f, i) Fluent f becomes true in interval i

18
Events

• We assume a distinguished event, Start , that describes the initial state

by saying which fluents are initiated or terminated at the start time.
• We define T by saying that a fluent holds at a point in time if the fluent
was initiated by an event at some time in the past and was not made
false (clipped) by an intervening event.
• A fluent does not hold if it was terminated by an event and not made
true (restored) by another event.
• Happens(e, (t1, t2)) ∧ Initiates(e, f, t1) ∧ ￢ Clipped(f, (t1, t)) ∧ t1 < t ⇒ T(f, t)
• Happens(e, (t1, t2)) ∧ Terminates(e, f, t1)∧ ￢ Restored (f, (t1, t)) ∧ t1 < t ⇒ ￢
T(f, t)
where Clipped and Restored are defined by
• Clipped(f, (t1, t2)) ⇔ ∃ e, t, t3 Happens(e, (t, t3)) ∧ t1 ≤ t < t2 ∧ Terminates(e, f,
t)
• Restored (f, (t1, t2)) ⇔ ∃ e, t, t3 Happens(e, (t, t3)) ∧ t1 ≤ t < t2 ∧ Initiates(e, f, t)

19
Processes

• The events we have seen so far are what we call discrete events
• Categories of events with sub-intervals are called process categories
or liquid event categories

20
Time intervals

 Time intervals are partitioned into moments (zero

duration) and extended intervals

Partition(Moments,ExtendedIntervals,Intervals)
i i Intervals  (i  Moments  Duration(i)  0)
 Functions Start and End delimit intervals
i Interval(i)  Duration(i)  (Time(End (i))  Time(Start
(i)))
 May use e.g. January 1, 1900 as arbitrary time 0
Time(Start(AD1900))=Seconds(0)

21
Relations between time intervals

Meet(i, j) j

Before(i, j) i
After(j, i) j

i Can be
j expressed
During(i, j )
logically,
i e.g.
Overlap(i, j)
i, j Meet (i, j)
Overlap(j, i) j
 Time(End (i))
 Time(Start ( j))
22
Mental events and mental objects

 Need to represent beliefs in self and other agents, e.g.

for controlling reasoning, or for planning actions that
involve others
 How are beliefs represented?
•  Beliefs are reified as mental objects
•  Mental objects are represented as strings in a
language
•  Inference rules for this language can be defined
 Rules for reasoning about logical agents’ use their
beliefs
• a, p,q LogicalAgent (a)  Believes(a, p) 
a, p LogicalAgent (a)  Believes(a, p)
• Believes(a," p  q" )  Believes(a,q)
 Believes(a," Believes( Name(a), p)" )

23
Mental events

• propositional attitudes that an agent can have toward mental

objects: attitudes such as Believes, Knows, Wants, Intends, and
Informs
• For example, suppose we try to assert that Lois knows that Superman
can fly:
Knows(Lois, CanFly(Superman))
• if it is true that Superman is Clark Kent, then we must conclude that
Lois knows that Clark can fly:
(Superman = Clark) ∧ Knows(Lois , CanFly(Superman)) |= Knows(Lois,
CanFly(Clark ))
• This property is called referential transparency

24
Modal Logic

• Modal logic is designed to address this problem.

• Regular logic is concerned with a single modality, the modality of
truth, allowing us to express “P is true.”
• Modal logic includes special modal operators that take sentences
(rather than terms) as arguments.
• For example, “A knows P” is represented with the notation KAP,
where K is the modal operator for knowledge. It takes two
arguments, an agent (written as the subscript) and a sentence.

25
 Semantic networks

 Graph representation of categories, objects,

relations, etc. (i.e. essentially FOL)

 Natural
representation
of inheritance
and default
values
∀x x∈ Persons ⇒ [∀ y
HasMother (x, y) ⇒
y ∈ FemalePersons ] .

∀x x∈ Persons ⇒ Legs(x, 2) . 26
Semantic Network

Is a Human
Boy Being
a
a Is
Is
Goes to Needs
Woman
School Joe
a
Is
Has Food
a child
Kay

27
Other reasoning systems for categories

 Description logics
 Derived from semantic networks, but more
formal
 Supports subsumption, classification and
consistency
 Circumscription and default logic
 Formalizes reasoning about default values
 Assumes default in absence of other input; must
be
able to retract assumption if new evidence occurs
 Truth maintenance systems
 Supports belief revision in systems where
retracting belief is permitted
28
Internet shopping world

 An agent that understands and acts in an

internet shopping environment
 The task is to shop for a product on the Web,
given the user’s product description
 The product description may be precise, in which
case the agent should find the best price
 In other cases the description is only partial, and
the agent has to compare products
 The shopping agent depends on having product
knowledge, incl. category hierarchies

29
PEAS specification of shopping agent

 Performance goal
 Recommend product(s) to match user’s
description
 Environment
 All of the Web
 Actions
 Following links
 Retrieve page contents
 Sensors
 Web pages: HTML, XML

30
Outline of agent behavior

 Start at home page of known web store(s)

 Must have knowledge of relevant web
addresses,
such as www.amazon.com etc.
 Spread out from home page, following links to
relevant pages containing product offers
 Must be able to identify page relevance, using
product category ontologies, as well parse
page contents to detect product offers
 Having located one or more product
offers,
agent must compare and recommend
product
31
 Comparison range from simple price
Following links

• The agent will have knowledge of a number of stores, for example:

Amazon ∈OnlineStores ∧ Homepage(Amazon, “amazon.com”) .
Ebay ∈OnlineStores ∧ Homepage(Ebay, “ebay.com”) .
ExampleStore ∈OnlineStores ∧ Homepage(ExampleStore, “example.com”)
• a page is relevant to the query if it can be reached by a chain of zero or
more relevant category links from a store’s home page, and then from
one more link to the product offer.
Relevant(page, query) ⇔ ∃ store, home store ∈OnlineStores ∧
Homepage(store, home) ∧ ∃url , url 2 RelevantChain(home, url 2, query) ∧
Link(url 2, url ) ∧ page = Contents(url )

RelevantChain(start , end, query) ⇔ (start = end) ∨ (∃ u, text LinkText(start,

u, text ) ∧ RelevantCategoryName(query, text ) ∧ RelevantChain(u, end,
query)) . 32
Following Links

33
Comparing offers

∃ c, offer c∈ LaptopComputers ∧ offer ∈ ProductOffers ∧

Manufacturer(c,IBM ) ∧ Model (c, ThinkBook970 ) ∧
ScreenSize(c, Inches(14)) ∧ ScreenType(c, ColorLCD) ∧
MemorySize(c,Gigabytes(2)) ∧ CPUSpeed (c,GHz (1.2)) ∧
OfferedProduct(offer, c) ∧ Store(offer , GenStore) ∧
URL(offer , “example.com/computers/34356.html”) ∧
Price(offer , $(399)) ∧ Date(offer ,Today)

34
Summary

 An ontology is an encoding of vocabulary and

relationships. Special-purpose ontologies can
be effective within limited domains
 A general-purpose ontology needs to cover a wide
variety of knowledge, and is based on categories
and an event calculus
 It covers structured objects, time and space,
change, processes, substances, and beliefs
 The general ontology can support agent
reasoning in a wide variety of domains, including
the Internet shopping world

35