Rules for Transforming OWL 2 Ontology into SBVR

Gintare Krisciuniene

, Lina Nemuraite

, Rita Butkiene

and Bronius Paradauskas

Departament of Information Systems, Kaunas University of Technology, Studentu st.50-309, Kaunas, Lithuania

Centre of Information Systems Design Technologies,

Kaunas University of Technology, Studentu st.50-313, Kaunas, Lithuania

Keywords: Ontology, Business Concepts, Business Rules, Representation, Meaning, SBVR, OWL 2, Transformation.

Abstract: Our research is concentrated on defining transformation rules from OWL 2 ontologies into SBVR

vocabularies and rules without a loss of information and the expressive power, characteristic for ontologies,

overcoming the fact that some ontology-specific concepts have no direct representation in SBVR. Our focus

is on generic transformation rules, but the particular attention is devoted to ontologies and vocabularies

related with semantic search in Lithuanian Internet corpus. Therefore, we consider some particular

constructs related with our application domain, including the idea of creating domain-specific lexical

ontologies, related with domain ontologies and capable to support semantic annotating and search.

1 INTRODUCTION

Web Ontology Language OWL 2 (W3C, 2012) is a

knowledge representation language, used for sharing

a common understanding of a certain domain among

computer systems and human experts, and having

capabilities for reasoning and querying semantic

specifications. OWL 2 is not easily understood by

every user. Semantics of Business Vocabulary and

Business Rules (SBVR) (OMG, 2013) provides

opportunity to describe business concepts and

business rules in the structured language, similar to

natural language and understandable for human.

SBVR is based on formal logics and can be applied

for computer processing, but it cannot be directly

used in semantic technologies.

Both languages, OWL 2 and SBVR, are created

for expressing semantics of the domain, but the

development of these languages was inspired by

different issues. In result, we have two different

metamodels and different sets of tools for semantic

processing. The semantically overlapping concepts

of both knowledge models has encouraged

investigating a possibility of transforming SBVR

into OWL 2 and vice versa. The transformation of

SBVR to OWL 2 will allow business users to

describe domain ontologies using human-

understandable language, to prove consistency of

business vocabularies and rules by using OWL 2

reasoners, etc. The reverse transformation will allow

business users to have a human-friendly interface to

ontologies, considering them in business

applications (Ghali, 2012) and in semantic search,

where SBVR questions in structured language are

transformed into SPARQL queries (Sukys, 2012a;

2012b). SBVR to OWL 2 transformation was

considered in several works, of which the

transformation of (Karpovic et al., 2011; 2012)

seems the most comprehensive and suitable for

further investigation.

The goal of current paper is to present rules for

transforming OWL 2 ontologies into SBVR,

compatible with SBVR to OWL 2 transformation

(Karpovic, 2012). Both these transformations should

be mutually reversible and lossless, and compatible

with SBVR to SPARQL transformation, because in

semantic search SBVR questions are transformed

into SPARQL queries, which are executed in the

source ontology (Sukys, 2012a).

Semantic search in our research is directed

towards better facilitation of people and

organizations to use natural Lithuanian language in

the virtual space in their professional and personal

activities. Natural language technologies require

sophisticated processing algorithms and vast

amounts of resources. In the world practice, lexical

resources as WordNet, VerbNet, FrameNet, or

PropBank are used for relating senses of ontology

elements with their verbal representations in

semantic annotating and search. The representation

256

Krisciuniene G., Nemuraite L., Butkiene R. and Paradauskas B..

Rules for Transforming OWL 2 Ontology into SBVR.

DOI: 10.5220/0005076902560263

In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2014), pages 256-263

ISBN: 978-989-758-049-9

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

part of SBVR is similar to what is encompassed by

WordNet and other lexical ontologies where various

syntactic forms are related to meaning. We had

made an assumption (Bernotaityte, 2013;

Krisciuniene, 2014) that the lexical ontology for

Lithuanian language can be based on SBVR

representations and related to SBVR based domain

ontology thus making it possible to accelerate the

task of creating the lexical resources required for

embodying the semantic search techniques for

Lithuanian language.

The rest of the paper is structured as follows.

Section 2 analyses related work. Section 3 presents a

domain ontology example. Section 4 describes rules

for transforming OWL 2 ontologies into SBVR.

Section 5 presents initial experiments for checking

their applicability. Section 6 summarizes

conclusions and envisages future research.

2 RELATED WORKS

Currently, there are several research works and

prototypes aiming at transforming SBVR business

vocabulary and business rules into ontologies

(ONTORule, 2009; Karpovic, 2011, 2012; Kendall,

2013; Reynares, 2013); some informal mappings

between SBVR and OWL 2 concepts are given in

SBVR specification (OMG, 2013). The

transformations described in (Karpovic, 2011, 2012)

are the most appropriate for our purpose as they are

not only comprehensive, bet also take into account

aspects specific for our joint research. SBVR to

OWL 2 transformations often are defined from the

SBVR side: what constructs of SBVR can be

represented in OWL 2. Our research needs not only

making the SBVR  OWL 2 transformations

mutually reversible but also to reflect ontology

advantages in SBVR specifications.

The result of SBVR to OWL 2 transformation

(Karpovic, 2011, 2012) is the domain ontology

based on preferred representations of SBVR

concepts. This transformation does not involve

synonyms and synonymous forms except a single

synonymous form for each verb concept wording, if

it is required for specifying business rules and

obtaining the corresponding inverse object property.

SBVR metamodel does not give possibility for

specifying desirable inverse verb concept wordings.

For solving this and other similar problems,

additional concept types (e.g., inverse verb concept)

are described in the special SBVR for OWL 2

vocabulary, which can be incorporated into SBVR

vocabularies for transforming them into OWL 2.

(Kendall and Linehan, 2013) define the solid

reversible SBVR to OWL 2 transformation without

loss of semantic information but the result of reverse

transformation does not guarantee an identical

original representation, because they transform

SBVR synonyms and synonymous forms into

OWL 2 annotations. The reverse transformation

from OWL 2 into SBVR is not capable to recover

the original vocabulary and rules although they

remain semantically equivalent. For solving this

problem, we propose separating SBVR synonyms

and synonymous forms from the domain ontology

and creating the lexical ontology based on SBVR

representations (Krisciuniene, 2014). Other analysed

SBVR to OWL 2 transformations were superficial,

e.g., (Reynares, 2013), but the author’s idea for

preserving information about SBVR partitive verb

concepts in ontology is noteworthy, and we are

willing to borrow it with the reference to the authors.

There was found only one (except our) work dealing

with transformation from ontologies to SBVR

(Gailly, 2013), however, it was in an initial stage

and did not propose some special ideas or

experience.

The novelty of our work is that we are aiming at

creating SBVR vocabularies by building them upon

existing or designed ontologies, and considering

creating lexical ontologies, based on SBVR

representations, for the complete correspondence

and transformation between the transformable

subsets of SBVR and OWL 2 elements. Also, we use

a few SBVR and OWL 2 extensions for obtaining

the reversible and lossless OWL 2 – SBVR

transformations.

3 EXAMPLE OF ONTOLOGY

FOR SEMANTIC SEARCH

The excerpt from the OWL 2 ontology for Semantic

Search in Lithuanian Internet Corpus is presented in

Figure 1. The ontology is built on the base of SBVR

knowledge model, which presents the metamodel

and principles for creating domain specific

conceptual schemas (i.e., domain ontologies). The

top construct is the abstract “object”, which may be

any concept, having occurrence in the corpus under

investigation. More specific concepts are the agent,

location, time and “state of affairs”, which,

following the SBVR, may be an event, state,

activity, circumstance, etc. We focus on the generic

model of events, which may be related with an

agent, time, location, target and the event type that

RulesforTransformingOWL2OntologyintoSBVR

257

Figure 1: Excerpt from domain ontology for “saying” events (represented as UML class diagram).

allows the multiple categorizations of events, which

sometimes is

required instead of assigning the

unique category. The event is the problematic

concept. In a natural language, events are usually

expressed by verbs; majority of events represent n-

ary relations. SBVR allows describing n-ary

relations; however, the OWL 2 is limited to binary

ones. For describing n-ary relations in OWL 2, we

define the event class (SBVR general concept)

objectifying the n-ary relation (SBVR verb concept)

and having n relations with other classes (agent,

time, location, etc.). The number of roles may vary

depending on the type of the event and on

completeness of information we have. In our

research, SBVR models are limited to binary

relations due to difficulties of representing such

relations in software models and implementing in

CASE tools, SBVR and ontology editors, etc.

There are many ontologies and research works

devoted to event models, e.g., (Kaneiwa, 2007),

(Scherp, 2009), which describe temporal, spatial,

instance, participation, causality, mereology,

correlation, documentation, interpretation and other

event categories. The event (as well as other objects)

may be identified in many Web documents and is

important for semantic search. We concentrate on

“saying” event, as it is one of the most frequently

occurring events in Lithuanian Internet corpus. It is

related with speech acts, defined by Winograd and

Flores, which mean actions expressed by saying,

e.g., obligation, confirmation, agreement, etc.

Therefore, we suppose that “saying” is worth for the

primary attention in semantic annotating and search,

and, therefore, should be analysed first.

4 ONTOLOGY CONCEPTS AND

RULES FOR TRANSFORMING

THEM INTO SBVR

Main concepts of the OWL 2 are axioms and entities

(classes, object properties, data properties, data types

and named individuals) (Figure 2). Main SBVR

concepts are presented in Figure 3. For SBVR

structured language specifications, we use the SBVR

style of

terms, verbs, Names and keywords (OMG,

2013), where terms represent noun concepts (general

concepts, roles and verb concept roles); verbs

represent symbols used in verb concept wordings

(meaning verb concepts) and in statements meaning

propositions (facts). Vocabulary entries introduce

the primary forms (preferred representations) of

SBVR concepts, and can have captioned details,

e.g., General concept, Concept type, Synonym or

Synonymous form, etc. SBVR vocabulary can be

obtained from the OWL 2 ontology, and defined as

the individual concept of the general concept

“vocabulary”. Vocabulary name, namespace and

language can be obtained from the ontology name,

namespace and language. OWL 2 Annotations can

be used to specify additional information in

ontology, e.g., comments. We use standard

annotation property “

label” for human readability

KEOD2014-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

258

of entity names in a vocabulary language (e.g.,

English, as in the following example, or Lithuanian),

and the additional annotation properties, e.g.,

“

vocabularyURI”; “label_sbvr”, etc., for specifying

entity names in SBVR style that would be useful for

transforming OWL 2 ontology into SBVR.

OWL 2 Entities define named elements of the

OWL 2 ontology, uniquely identified by their IRIs

and declared by the Declaration axioms. OWL 2

Class is transformed into SBVR general concept,

ObjectProperty with its domain and range – into

verb concept; OWL 2 DataProperty corresponds to

the SBVR role, except DataProperty with

DataRange ‘boolean’, which corresponds to SBVR

unary verb concept (characteristic).

OWL 2 has a rich set of data types including RDFS

Literals, RDF DataTypes, XSD DataTypes and Plain

Literals (W3C, 2012). SBVR has just a few

elementary concepts (text, URI, number, integer,

nonnegative integer, positive integer) that can be

used for representing the corresponding OWL 2 data

types. However, the SBVR allows extensions.

SBVR extension for Data and Time (OMG, 2011)

defines various extensions of SBVR elementary

concepts for representing dates and time durations.

For the OWL 2  SBVR transformation, we have

introduced into SBVR

boolean and date_time as the

most necessary elementary concepts. Further

extensions can be added as necessary.

Figure 2: Main concepts of OWL 2 (some class axioms are shown in more detail).

Figure 3: Main concepts of SBVR meaning extended with concepts for reflecting advanced features of OWL 2.

RulesforTransformingOWL2OntologyintoSBVR

259

OWL 2 Class Axioms and Class Expressions.

SubClassOf class axiom provides possibility to create

class specialization hierarchies by defining the

subsumption dependency between classes. Such

axiom can be transformed to SBVR categorization:

OWL: SubClassOf(saying event) 

SBVR: event

saying

General concept: event

OWL 2 SubclassOf axioms are formulated along

with many other OWL 2 axioms and restrictions:

AllValuesFrom, SomeValuesFrom, ObjectHasSelf,

ObjectHasValue, cardinality restrictions, etc.; in such

cases,

SubclassOf axioms are not transformed.

Class expression

AllValuesFrom, defining

universal quantifications on object properties or data

properties, can be transformed to SBVR necessity

statements scoping over universal quantifications

and atomic formulations, which use more general

verb concepts for defining specific general concepts

as players of more general roles:

OWL: SubClassOf(journey (ObjectAllValuesFrom

(has__target_object location)) 

SBVR: It is necessary that journey

has_target_object that is location

Class expressions SomeValuesFrom, defining

existential quantifiers on object properties or data

properties, are transformed to SBVR necessity

statements scoping over existential quantifications

and atomic formulations based on binary verb

concepts, e.g.:

OWL: SubClassOf(saying ObjectSomeValuesFrom

(is_said_about saying_object))

SBVR: It is necessary that saying is_said_about

saying_object at least 1 saying_object

OWL 2 EquivalentClasses axiom denotes the

equivalence of class expressions. This axiom

between single classes is transformed into SBVR

verb concept

concept1 is_coextensive_with

concept2. The axiom between a single class and a

class expression, which defines how this class is

derived, is transformed to the SBVR definition.

Class disjointness in the OWL 2 means that an

individual

I can be an instance of the only one class

(class expression)

CEi from the set of disjoint

classes.

DisjointClasses can be transformed to

SBVR impossibility statements, or necessity

statements with nor formulations.

DisjointUnion (C,CE

, …, CE

), n≥2, states that a

class

C is the disjoint union of classes CE

, …, CE

which are pairwise disjoint.

DisjointUnionOf axiom

can be transformed to SBVR disjunction

accompanied with impossibility statement or nor

formulation, e.g.:

OWL: SubClassOf(person, agent)

SubClassOf(organization, agent)

DisjointClasses(person, organization) 

SBVR: agent

Definition: person or organization

person

General concept: agent

organization

General concept: agent

It is impossible that person is organization

The ObjectUnionOf, ObjectIntersectionOf, and

ObjectComplementOf class expressions can be

transformed to SBVR logical operations with the

closed logical formulations, e.g.:

OWL: EquivalentClasses(person

ObjectIntersectionOf (ObjectComplementOf

(organization) agent)) 

SBVR: person

Definition: agent that is not organization

The OWL 2 ObjectHasValue class expression

allows expressing object properties of individuals. In

the SBVR, such expression can be specified as a

fact, based on the verb concept, in which one role is

played by an individual verb concept.

OWL 2 ObjectPropertyExpressions and

ObjectPropertyAxioms.

OWL 2

InverseObjectProperties axiom denotes

that two object properties

and OP

are pair-wise

inverse, e.g.:

OWL: InverseObjectProperties(said__saying

is_said_by__speaker)

SBVR: speaker said saying

Synonymous_form: saying said_by speaker

saying said_by speaker

Concept type: inverse_verb_concept

Transformation rules for the

SubObjectProperty,

DisjointObject Properties,

EquivalentObjectProperties

axioms are similar to

the transformation rules of

SubclassOf,

DisjointClasses and EquivalentClasses axioms.

ObjectPropertyChain is the more complex

SubObjectPropertyAxiom. It states that if an

individual

is connected by a chain of object

property expressions with an individual

, then I

also connected with

by the derivable object

property expression

OPE. ObjectPropertyChain can

be transformed to the SBVR necessity statement

formulated by the implication formulation, which

has the antecedent restricted by one or more

projecting formulations, and the second role of its

consequent coincides with the second verb concept

role of the last verb concept in the projecting

formulations’ chain, e.g.:

KEOD2014-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

260

OWL: SubObjectPropertyOf(ObjectPropertyChain

(said__saying is_part_of__event)

is_participant_of__event))



SBVR: It is necessary that agent

is_participant_of event if agent said

saying that is_part_of event

ObjectHasSelf axiom allows specifying the

object property that is pure reflexive in ORM2 terms

(Halpin, 2005).

FunctionalObjectProperty can be transformed to

the SBVR

at_most_one_quantification, e.g.:

OWL:FunctionalObjectProperty(occurs_in_location)

SBVR: It is necessary that

event occurs_in

at most 1 location

The Inverse Functional Property cannot be

directly specified in SBVR. Besides this, there is a

set of OWL 2 Object Property Axioms that come

from ORM2 (Halpin, 2005) and are important for

inference:

Reflexive, Irreflexive, Symmetric,

Assymetric, and Transitive object properties that do

not have corresponding characteristics in SBVR

metamodel, though the latter also is based on ORM2

(Halpin, 2011). For solving this problem, we

extended the SBVR binary verb concept similarly as

in the case of inverse object property (Figure 3).

Transformation of characteristics of OWL 2 object

properties to concept types of SBVR verb concepts

is straightforward, e.g.:

OWL: TransitiveObjectProperty

(is_part_of__event) 

SBVR:

event_part is_part_of event

Concept type: transitive_verb_concept

Transformations of OWL 2 Cardinality

Restrictions also are straightforward, as they have

their direct equivalents in SBVR, except

DataExactCardinality ‘1’ on Data Properties having

DataRange ‘boolean’. These restrictions are

transformed to metamodel level statements of the

type

‘concept incorporates characteristic’, e.g.:

OWL: SubClassOf(object (DataExactCardinality

(1 is_trusted xsd:boolean))

SBVR: object incorporates characteristic

‘is_trusted’

Transformation rules for OWL 2 DataProperty

axioms and restrictions,

ObjectPropertyAssertions

and

DataPropertyAssertions are defined in a similar

way. The short summary of OWL 2 to SBVR

transformation rules is given in Table 1.

5 EXPERIMENTAL APPROVAL

Two experiments were conducted for transforming

OWL 2 ontologies into SBVR vocabularies. For

evaluating the suitability of defined transformations

for existing ontologies, the OWL2SBVR prototype

has been implemented using ASP.NET technologies

(the final implementation is under development in

ATL transformation language as an integral part of

the overall framework for the semantic search).

Also, the purpose of the experiment was to find the

unexpected problems, which could remain unknown

from the context of theoretical models.

For the first experiment, three ontologies from

internet (SIOC, Wine and GoodRelations) were

chosen and used without any preparation for

obtaining comprehensible vocabularies. Names of

vocabulary entries were obtained from labels (or

names, if labels were missing) of ontology entities.

Names of classes, individuals and data properties,

consisting of several words, were reconstructed into

SBVR style (in ontologies, they usually are

constructed using camel style). Names of verb

concepts were constructed from names of domain

classes, object properties and range classes.

The experiment has shown that almost all simple

elements (classes, properties, class hierarchies) can

be transferred to the SBVR vocabulary. All classes

of SIOC, Wine and GoodRelations were transferred.

The problems have arisen with transformation of

properties. In the worst case, only 48.2% of object

properties and 70% of data properties were

transformed from SIOC ontology. The semantics of

properties was often questionable, and could not be

automatically ensured.

Several problems were identified during the first

experiment. Object properties sometimes have no

domain and range specified. In such cases, the

“thing” class may be assigned as a domain or range

by default, but it would not be a right solution in all

cases. Object properties often are named by nouns

that have meaning of roles; or they present junctions

of relations and roles, which are expressed by

phrases, consisting from several words. Camel style

does not help to recognize such constructions.

Moreover, object properties can have several

domains or ranges, or have excess information, as in

the labels (e.g., "type of good (1..1)"@en)) of Good

Relations ontology.

In general, there is no rational way to

automatically recognize roles in OWL 2 ontologies.

Other problems have arisen with multiple

categorizations when the same concept belongs to

several subsumption hierarchies. While SBVR

RulesforTransformingOWL2OntologyintoSBVR

261

Table 1: The short summary of rules for transforming OWL 2 constructs into SBVR.

OWL 2 Entities

OWL 2 SBVR

EntityIRI NameSpace URI

Class general_concept

NamedIndividual individual_concept

ClassAssertion classification

string, boolean, integer, nonnegative integer,

possitiveInteger , dateTime

text, boolean, integer, nonnegative integer, positive_integer,

number, date_time

ObjectProperty association

O bjectProperty, SubObjectPropertyOf ( ObjectProperty

Partitive_object_property)

partitive_verb_concept

DataProperty, DataPropertyDomain, DataPropertyRange

property_association | characteristic

OWL 2 Class expressions and axioms

SubClassOf categorization

ObjectAllValuesFrom|DataAllValuesFrom

necessity_statement with universal_quantification

ObjectSomeValuesFrom]DataSomeValuesFrom

necessity_statement with existential quantification

SubClassOf, DataSomeValuesFrom for DataRange

‘boolean’

concept incorporates characteristic

EquivalentClasses for single classes

association ‘ concept

is_coextensive_with concept

’

EquivalentClasses between a class and axiom

definition

EquivalentClasses, SubClassOf hierarchy,

DisjointUnion

segmentation

EquivalentClasses, SubClassOf hierarchy ccategorization_scheme

DisjointClasses

impossibility_statement| necessity_statement with

nor_formulation

ObjectUnionOf, ObjectIntersectionOf,

ObjectComplementOf, ObjectOneOf, ObjectHasSelf,

ObjectHasValue

disjunction, conjunction, logical_negation, definition with

conjunction of Individual concepts,

purely_reflexive_verb_concept fact with Individual concept

CardinalityRstriction

quantification

OWL 2 Property Expressions and Axioms

ObjectProperty, InverseObjectProperties verb_concept, inverse_verb_concept

SubObjectProperty, SubDataProperty

categorization of verb concepts

DisjointObjectProperties, DisjointDataProperties

impossibility_statement for verb concepts

EquivalentObjectProperties, EquivalentDataProperties association ‘ concept

is_coextensive_with concept

’

ObjectPropertyChain

necessity_statement with implication_formulation and

projecting_formulation chain

FunctionalObjectProperty, FunctionalDataProperty at_most_one_quantification

InverseFunctionalObjectProperty

at_most_one_quantification for inverse_verb_concept

(Transitive|Reflexive|Irreflexive|Symmetric|Assymetri

c ObjectProperty)

(transitive|reflexive|irreflexive|symmetric|assymetric

verb_concept)

supports multiple categorizations, such cases should

have a special structure leading to categorization

schemes or segmentations. According ontology

normalisation requirements, multiple categorizations

are allowable in OWL 2 ontologies in the form of

derivable classes. Existing ontologies may not fulfil

such requirements.

The second experiment was performed with

ontologies of three domains related with semantic

search (politics; business and economy, and public

administration). These ontologies were developed in

accordance with the rules of the normalization and

other requirements for ontologies intended to be

used for creating SBVR vocabularies. These

ontologies had dedicated annotations “label_sbvr”,

specified using SBVR style in the Lithuanian

language.

The experiment has shown, that it is possible to

correctly reflect the semantics of ontologies via

concepts of SBVR business vocabulary and business

rules if these ontologies follow the normalisation

rules and are provided with the desired

representations, especially for specifying verb

concept roles in ontology properties. Existing

ontologies can be prepared for creating SBVR

vocabularies in any language as in our framework

for semantic search transformations are language-

independent.

KEOD2014-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment

262

6 CONCLUSIONS AND FUTURE

WORKS

The paper presents the rules for transforming

OWL 2 ontologies into SBVR business vocabularies

and business rules, which are intended for using

interlinked SBVR vocabularies and ontologies in

semantic search or other business applications.

Particularly, we are interested in semantic search in

Lithuanian Internet corpus; therefore, ontologies

reused or developed for that purpose should be

extended with specific labels allowing specifying

Lithuanian words and word phrases for naming

entities of the domain ontologies in the spoken

language and the style of SBVR. The experiments

have shown that freely chosen ontologies could

require some preparation before transforming them

to SBVR vocabularies: providing special labels,

ensuring ontology normalisation, supplementing

them with semantics of part-whole relations, etc.

The performed analysis has inspired extensions

of SBVR required for transforming inverse object

properties and characteristics of object properties. It

still remains a problem to transform OWL 2 object

properties without domains and ranges specified.

Sometimes, domains and ranges may be inferred

from property subsumption hierarchies or inverse

object properties. Also, we yet have not considered

complex domain and range specifications and other

advanced features that would require additional

efforts as well as the wider experimental

investigation of the proposed transformations.

ACKNOWLEDGEMENTS

The work is supported by the project VP1-3.1-

ŠMM-10-V-02-008 “Integration of Business

Processes and Business Rules on the Basis of

Business Semantics” (20132015), which is funded

by the European Social Fund (ESF).

REFERENCES

Bernotaityte, G., Nemuraite, L., Butkiene, R.,

Paradauskas, B., 2013. Developing SBVR

vocabularies and business rules from OWL2

ontologies. In ICIST 2013, CCIS, vol. 319. Springer,

Heidelberg, 134145.

Gailly, F., 2013. Transforming Enterprise Ontologies into

SBVR formalizations. In Proceedings of the CAiSE'13

Forum. Valencia, Spain, pp. 122129.

Ghali, A.E., Chniti, A., Citeau, H., 2012. Bringing OWL

ontologies to the Business Rules Users, In RuleML

2012 Proceedings. Montpellier, pp. 6276.

Halpin, T., 2005. ORM 2 Graphical Notation. Neumont

University, Technical Report ORM2-01.

Halpin, T., Curland, M., 2011. Enriched Support for Ring

Constraints. In On the Move to Meaningful Internet

Systems: OTM 2011 Workshops. LNCS, 7046.

Springer: Berlin, Heidelberg, pp. 309318.

Kaneiwa, K., Iwazyme, M., Fukuda, K., 2007. An upper

ontology for event classifications and relations. In AI

2007. LNAI 4830. Springer, Heidelberg, pp. 394-403.

Karpovic, J., Nemuraite, L., 2011. Transforming SBVR

business semantics into Web ontology language

OWL2: main concepts. In Information Technologies'

2011. Kaunas, Lithuania, pp. 231−238.

Karpovic, J., Nemuraite, L., Stankeviciene, M., 2012.

Requirements for Semantic Business Vocabularies and

Rules for Transforming them into Consistent OWL2

Ontologies. In ICIST 2012, CCIS, vol. 319. Springer,

Heidelberg, pp. 420435.

Kendall, E., Linehan, M. H., 2013. Mapping SBVR to

OWL2. IBM Research Report, RC25363 (WAT1303-

040).

Krisciuniene, G., Nemuraite, L., Butkiene, R.,

Paradauskas, B., 2014. Lexical Ontology Model Based

on SBVR Representations. In Computing,

Communication and Information Technology (CCIT

2014), London, UK, pp.77-81

OMG, 2011. Date-Time Vocabulary (Date-Time). Version

1.0. OMG Document bmi/2011-08-01.

OMG, 2013. Semantics of Business Vocabulary and

Business Rules (SBVR). Version 1.2. OMG Document

Number: formal/2013-11-05.

ONTORule Project, 2009. ONTOlogies meet Business

RULEs. ILOG/IBM, http://ontorule-project.eu/.

Reynares, E., Caliusco, M.L., Galli, M.R., 2013. An

Automatable Approach for SBVR to OWL 2

Mappings. In XVI Ibero-American Conference on

Software Engineering CIbSE 2013. Montevideo,

Uruguay,

pp. 201214.

Scherp, A., Franz, T., Saathoff, C., Staab, S., 2009. F-A

Model of Events based on the Foundational Ontology

DOLCE+DnS Ultralite. In K-CAP 2009. California,

pp. 137144.

Sukys, A., Nemuraite, L., Paradauskas, B., Sinkevičius,

E., 2012a. Transformation framework for SBVR based

semantic queries in business information systems. In

Bustech 2012. Nice, France, IARIA, pp. 16.

Sukys, A., Nemuraite, L., Paradauskas, B., 2012b.

Representing and transforming SBVR question

patterns into SPARQL. In ICIST 2012, CCIS 319.

Springer, Heidelberg, pp. 436451.

W3C, 2012. OWL 2 Web Ontology Language Structural

Specification and Functional-Style Syntax (Second

Edition). W3C Recommendation 11 December 2012.

RulesforTransformingOWL2OntologyintoSBVR

263