Transcript Chronos: A Tool for Handling Temporal Ontologies in Protege

Chronos: A Tool for Handling Temporal
Ontologies in Protégé
Alexandros Preventis, Polyxeni Marki,
Euripides G.M. Petrakis, Sotirios Batsakis
Dept. of Electronic and Computer Engineering
Technical University of Crete (TUC)
1
OWL
 OWL (Web Ontology Language) is a language that can be
used to describe the classes and relations between them.
 more expressive than XML, RDF and RDF-S
 allows to reason about the entities and check whether or not
all statements and definitions are mutually consistent
 OWL directly supports binary relations.
2
Dynamic Ontologies
 Represent concepts that occur and evolve in time
 A company will be established, hire personnel and
develop products
 The relation “employs" and its inverse are ternary.
3
OWL-Time Ontology
 OWL-Time is an OWL ontology of temporal concepts. It
provides a vocabulary for describing:
 topological relations between temporal entities (instants,
intervals)
 information about durations
 It provides no means of representing information that changes
in time.
4
Temporal Representations
 Models for representing temporal information:
 Temporal Description Logics (TDLs)
 Versioning
 Reification
 4D-Fluents (Welty 2006 & Batsakis 2011)
 N-ary Relations
5
N-ary Example
6
Protégé
 Protégé is a free, open-source editor platform for handling
ontologies:
 enables users to build ontologies for the Semantic Web, in
particular in the W3C's Web Ontology Language (OWL)
 Can be extended by means of a plug-in architecture and a Javabased Application Programming Interface (API) for building
knowledge-based tools and applications
7
Problem
 The creation of a temporal ontology, or the conversion of a static




8
ontology to temporal can be a very complex, time-consuming and
error-prone procedure.
Involves intermediate relations and properties
The temporal properties in N-ary apply on intermediate relations
The user must have very good knowledge of the underlying (N-ary
in our work) representation model.
There is no tool for crafting temporal concepts in ontologies easily
(i.e., similarly to static ontologies).
Chronos
 We present Chronos, a Plug-In for the Protégé editor
(version 4.1,supporting OWL 2.0) that:
 handles temporal ontologies similarly to static ones
 supports restriction checking on temporal properties, classes
and individuals
 supports reasoning over temporal ontologies using the Pellet
reasoner with Protégé
 its interface is consistent with the layout of the default Protégé
tabs
 The user need not be familiar with the N-ary model
9
Modified entities
 Although it is feasible to create a temporal ontology using the
standard Protégé tabs, it is difficult to define temporal
properties and restrictions.
 Changes that are applied during the conversion of:
 Object Properties
 Data Properties
 Individuals
 Classes
10
Object Properties
 The changes that will take place in order to convert an object
property to temporal are:
 Property domain and range
 Event class is related to the Interval with the object property
during
11
Data properties
 The changes that will take place in order to convert a data
property to temporal are:
 Property domain
 creation of an object property (named by the data property, and
followed by “OP") which will relate the static data property's
domain to the Event class.
 Event class is related to the Interval with the object property
during
12
Individuals
 When a property is converted to temporal, all its instances
have to be converted as well.
13
Value constraints
 owl:allValuesFrom Company employs only Programmer In
the case where `employs' is temporal object property, the
constraint is:
 Company employs only (Event and (employs only
Programmer))
 owl:someValuesFrom Company employs some Programmer
In the case where `employs' is temporal object property, the
constraint is:
 Company employs some (Event and (employs some
Programmer))
14
Cardinality Constrains (1/2)
 owl:maxCardinality Company employs max 2 Employee
 owl:minCardinality Company employs min 2 Employee
 owl:cardinality Company employs exactly 2 Employee
 owl:FunctionalProperty A functional property is a property that can
have only one (unique) value y for each instance x, i.e. there cannot be
two distinct values y1 and y2 such that the pairs (x, y1) and (x, y2) are
both instances of this property.
 John worksFor Apple
 A temporal functional property can have only one value in each time
interval during which the property holds. This is implemented by adding
a SWRL rule to the ontology (Batsakis2012).
15
Cardinality Constrains(2/2)
 owl:inverseFunctionalProperty If we state that P is an
owl:InverseFunctionalProperty, then this asserts that a value y can only
be the value of P for a single instance x, i.e. there cannot be two distinct
instances x1 and x2 such that both pairs (x1, y) and (x2, y) are instances
of P.
 Apple employs John
 When a temporal property is inverse-functional the object uniquely
determines the subject for each time instant.
 If a temporal property P is transitive and (x, y, interval1) is an instance
of P and (y, z, interval2) is an instance of P, then we can infer that (x, z,
interval1 ⋂ interval2) is also an instance of P. In our implementation
the transitivity between instances of a temporal property takes place
only if those instances hold for same time intervals
16
Class Conversions
 Chronos provides a convenient way to convert multiple
object and data properties to temporal. This function
converts:
 the object and data properties that relate members of the
selected class
 the object and data properties where this class appears as a
Domain
 the restrictions where one of these object or data properties
appear in
17
User Interface
18
Conclusions
 Chronos, is a tab plug-in for Protege editor that facilitates the




19
creation and editing of temporal OWL 2.0 ontologies.
The temporal concepts as well as the properties that evolve over
time are represented by means of the N-ary Relations model, that
is a W3C recommendation.
It applies on restrictions on temporal properties, which have
different semantic meaning than those applied on static properties.
The user does not have to be familiar with the peculiarities of the
temporal representation model, thus making the manipulation of
temporal entities as easy as if they were static.
Available at: www.intelligence.tuc.gr/prototypes
Thank You
Questions?
20