Supporting Process Model Development
with Enterprise-Specific Ontologies
Nadejda Alkhaldi
1
, Sven Casteleyn
1,2
and Frederik Gailly
3
1
Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium
2
Universitat Jaume I, Av. De Vicent Sos Baynat s/n, 12071 Castellon, Spain
3
Ghent University, Tweekerkenstaart 2, 9000 Gent, Belgium
Keywords: Enterprise Modelling, Enterprise Ontology, Ontology-driven Modelling, BPMN.
Abstract: Within an enterprise, different models – even of the same type - are typically created by different modellers.
Those models use different terminology, are based on different semantics and are thus hard to integrate. A
possible solution is using an enterprise-specific ontology as a reference during model creation. This allows
basing all the models created within one enterprise upon a shared semantic repository, mitigating the need
for model integration and promoting interoperability. The challenge here is that the enterprise-specific
ontology can be very extensive, making it hard for the modeller to select the appropriate ontology concepts
to associate with model elements. In this paper we focus on process modelling, and develop a method that
uses four different matching mechanisms to suggest the most relevant enterprise-specific ontology concepts
to the modeller while he is creating the model. The first two utilize string and semantic matching techniques
(i.e., synonyms) to compare the BPMN construct’s label with enterprise-specific ontology concepts. The
other two exploit the formally defined grounding of the enterprise ontology in a core ontology to make
suggestions, based on the BPMN construct type and the relative position (in the model). We show how our
method leads to semantically annotated process models, and demonstrate it using an ontology in the
financial domain.
1 INTRODUCTION
When different models within an enterprise are
created by different modellers, integrating those
models is hard. This is known as the correspondence
problem (Fox and Grüninger, 1997). A possible
solution for this integration problem is providing
modellers with a shared vocabulary formalized in an
ontology (Bera et al., 2009). Over the last 30 years,
different domain ontologies have been developed
which describe the concepts, relations between
concepts and axioms of a specific domain. In a
business context, a particular type of domain
ontologies is so-called enterprise ontologies. They
describe the enterprise domain and consequently
provide enterprise domain concepts that can be
reused by different enterprises. Example of
enterprise ontologies include the Enterprise
Ontology (Uschold et al., 1998), TOVE (Fox, 1992)
and the Resource Event Agent enterprise ontology
(Gailly et al., 2008). Two different approaches have
been proposed to incorporate enterprise ontologies
into the modelling process. Some authors consider
enterprise ontologies to be reference models that
support the creation of different kind of models. For
instance, (Fox and Grüninger, 1997) suggests
developing the Generic Enterprise Model as an
ontology that is later used as a reference for creating
both data and process models. Other authors
developed an enterprise-specific modelling language
which is based on the concepts, relations and axioms
described in the enterprise ontology (Sonnenberg et
al., 2011). There are also works proposing a process
ontology to be used specifically for process models,
e.g., (Klein and Bernstein, 2001), (Schlenoff et al.,
1998) and (Haller et al., 2008). These works are
mainly concentrating on constructing those
ontologies, rather than supporting the modeller that
needs to use them.
In this paper we focus on using enterprise-
specific ontologies (ESO) during the development of
business process models. Enterprise-specific
ontologies are domain ontologies that differ from
enterprise ontologies in the fact that their Universe
236
Alkhaldi N., Casteleyn S. and Gailly F..
Supporting Process Model Development with Enterprise-Specific Ontologies .
DOI: 10.5220/0004899402360248
In Proceedings of the 16th International Conference on Enterprise Information Systems (ICEIS-2014), pages 236-248
ISBN: 978-989-758-029-1
Copyright
c
2014 SCITEPRESS (Science and Technology Publications, Lda.)
of Discourse is a specific enterprise, rather than the
enterprise domain. They may have their origin in an
established domain ontology or in an enterprise
ontology, but their main goal is describing the
concepts, relations and axioms that are shared within
a particular enterprise. Enterprise-specific ontologies
are getting increasingly important in the context of
Data Governance and knowledge representation
(Bera et al., 2011). Supporting tools, such as IBM
InfoSphere
1
or Collibra Enterprise Glossary
2
allow
enterprises to specify their own enterprise
glossary/ontology. Such an enterprise-specific
ontology, once available, can subsequently be
deployed to help enterprise modellers in creating
compatible, enterprise-specific models, such as
requirements, data or process models. This paper
focuses on business process models, and presents
mechanisms for assisting the business process
modeller by suggesting relevant terms while he is
constructing models, based on the enterprise-specific
ontology. Indeed, enterprise workers sometimes do
not know which knowledge they need to perform
their work (Bera et al., 2011), do not know all the
concepts they need to make their models and/or do
not know the agreed upon terminology. By
providing a limited set of relevant concepts to the
modeller, originating from the enterprise-specific
ontology, we effectively help him in his modelling
task, while promoting cross-model reuse of existing
concepts. The use of the enterprise-specific ontology
furthermore allows to semantically annotate
modelling elements in the resulting models, thereby
ensuring compatibility and integrateability of
different models.
The work described in this paper is part of a
larger framework for ontology-driven enterprise
modelling aimed to facilitate model construction
based on an enterprise-specific ontology on one
hand, and support enterprise-specific ontology
creation and evolution based on feedback from the
modelling process on the other hand. The framework
thus explores the symbiotic relationship between the
models and the ontology. This paper elaborates two
phases of the framework, namely the enterprise-
ontology-based model suggestion mechanisms, and
the model creation process. Other phases are not
elaborated here. To illustrate our work, we use the
Unified Foundational Ontology as core ontology,
OWL as ontology representation language and
BPMN as business process modelling language.
   
1
http://www-01.ibm.com/software/data/infosphere/
2
http://www.collibra.com/
2 ENTERPRISE-SPECIFIC
ONTOLOGY AND
MODELLING META-METHOD
As displayed in Figure 1, the previously outlined
framework (i.e., a meta-model) consists of two
parallel cycles: an ontology engineering cycle and
enterprise modelling cycle. Before a model can be
created, the enterprise-specific ontology needs to be
set up (Ontology Setup Phase in Figure 1). More
specifically, if not yet previously done, the concepts
and relationships contained in the enterprise-specific
ontology
3
(ESO) are analysed using a core ontology.
A core ontology is an ontology that describes
universally agreed upon, high level concepts and
relations, such as objects, events, agents, etc.
Grounding the enterprise-specific ontology in a core
ontology provides well-founded semantics, and
facilitates the modelling suggestion mechanisms (see
further). Once the enterprise-specific ontology is set
up, the enterprise modelling cycle starts with the
modeller selecting a model type (Enterprise Model
Selection phase), such as i*, ER, BPMN, Petri net,
etc. In our method, modelling languages that were
previously analysed using the same core ontology as
the ESO are particularly useful, as these facilitate
particular modelling suggestions based on the
enterprise-specific ontology, and using the core
ontology as an intermediary (see section 3). While
creating constructs in the model (Enterprise Model
Creation phase), aided by suggestions generated
based upon the ESO (Ontology Storage and
Suggestion Generation phase), the model is
automatically annotated with ESO concepts/relations,
and the modeller reports feedback on ontology
content: missing concepts or relations, or inaccurate
ones. When the model is finished, model quality is
evaluated (Enterprise Model Evaluation phase) and
once it satisfies certain quality standards, the
reported feedback is presented to the enterprise
stakeholders (i.e., analysts, modellers, developers,
managers, etc.) for discussion (Community-based
Ontology Feedback Evaluation phase). If this
community decides the feedback is valuable within
the context of the enterprise, it is added to
the enterprise-specific ontology (Ontology Evolution
phase), thus representing the evolution of the
ontology in a new version that corresponds better to
the business reality of the enterprise.
The proposed meta-method is unique because we
   
3
If no ESO is available, it can either be constructed from scratch,
or an existing domain ontology covering the business domain of
the enterprise may serve as a starting point.
SupportingProcessModelDevelopmentwithEnterprise-SpecificOntologies
237
Figure 1: Enterprise-specific ontology and modelling meta-method.
do not focus on a specific modelling language, and
we furthermore aim to exploit the symbiotic
relationship between the enterprise modelling cycle
and the ontology engineering cycle. Nevertheless,
several researchers have already performed research
isolated in a single phase of the meta-method.
Relevant for this article, several researchers have
investigated how business process models can be
semantically annotated using ontologies. (Liao et al.,
2013) describe how processes and subprocesses can
be annotaed by sets of ontology concepts. In (Lin
and Krogstie, 2010) the authors perform different
kinds of annotation on process models, including
profile, meta-model, model and goal annotation. Of
these annotations, model annotations are
immediately relevant for our work, which are stored
in a Process Semantic Annotation Model, along with
the elements of that model. In both these works, the
proposed method seems to be manual, and does not
focus on providing suggestions while the model is
under construction. Next to business process model
annotation, different researchers have also
investigated how natural Language Processing
techniques can be used while creating business
process models. For an overview we refer to
(Leopold, 2013). The focus here mainly lies with
extracting information and adding semantic
annotations to exiting models, rather than assisting
the modeller during modelling as in our work.
This paper focuses on the second phase of both
the ontology engineering and enterprise modelling
cycles. More specifically, it explains the
mechanisms that provide suggestion for the labels of
modelling construct suggestions to the modeller,
based on the enterprise-specific ontology, and details
how models are created and annotated with
enterprise-specific concepts.
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
238
3 SUGGESTION GENERATION
MECHANISMS
Our suggestion generation algorithm returns a
customized suggestion list of ESO concepts for (the
label of) every BPMN construct selected by the
modeller and placed on the canvas. Given the
potentially extensive amount of ESO concepts,
relevance ranking of suggestions is a critical feature.
Depending on the type of modelling construct that is
added, the position of the construct relevant to other
elements (i.e., its neighbourhood) in the model, and
the label entered by the modeller, the order of the
suggestion list is prioritized so that ontology
concepts with a higher likelihood to be relevant in
the current context appear first. To achieve this, four
different suggestion generation mechanisms are used
(Figure 2). These mechanisms are partly inspired by
ontology matching techniques, (Euzenat and
Shvaiko, 2013) but are specifically focused to fit
within our framework, where the semantics of the
modelling language (i.e., BPMN) meta-model can be
exploited. The first mechanism is the use of string
matching based on the BPMN element’s label
(partially or in full entered by the modeller). The
second is synonym matching, which retrieves
synonyms of the BPMN element’s label and verifies
whether there are syntactically equivalent concepts
in the ESO. The third mechanism derives
suggestions based on the relation between the
constructs of the modelling language (i.e. BPMN)
and the ESO concepts, using the core ontology (i.e.,
UFO) as intermediary. We call this “construct
matching”. The last suggestion generation
mechanism derives suggestions from the position of
a given BPMN construct relative to other elements
in the BPMN model. We call this “neighbourhood-
based matching”.
Every matching technique calculates a relevance
score (between 0 and 1) for each ESO concept,
which is stored. Subsequently, the overall relevance
score is calculated using a weighted average of all
individual scores. This corresponds to the formula
below:










Where:
: the score and weight of string match


: the score and weight of synonym match
: the score and weight of construct match
: the score and weight of neighbourhood based
match
The weights for each matching mechanism are
thus configurable. In our demonstration (see section
4), we assigned a higher weight to string matching
as we expect that, within a particular enterprise
context, a (quasi) exact string match has a high
possibility of representing the intended (semantic)
concept. The lowest weight is assigned to construct
matching, because it typically matches very broadly,
and thus delivers a large amount of suggestions.
Further experimentation with the distribution of
weights over the individual relevance scores should
be performed to determine an optimal overall score
calculation. This is considered future work. As a
final result, the suggestion list, a descending ordered
list of ESO concepts according to relevance, is
generated and presented to the modeller. In the next
four subsections the different mechanisms are
described in more detail. The implementation of the
mechanisms can be consulted via our Github
repository: https://github.ugent.be/MIS .
Figure 2: Suggestion generation algorithms.
SupportingProcessModelDevelopmentwithEnterprise-SpecificOntologies
239
final result, the suggestion list, a descending ordered
list of ESO concepts according to relevance, is
generated and presented to the modeller. In the next
four subsections the different mechanisms are
described in more detail. The implementation of the
mechanisms can be consulted via our Github
repository: [URL suppressed for blind review].
3.1 String Matching Mechanism
The goal of the string matching algorithm is to find
ESO concepts whose label is syntactically similar to
the label of BPMN elements entered by the modeller.
If these two strings are syntactically the same, there
is a high possibility that they have the same
semantics, especially as both reside within the same
enterprise and business context.
The string matching mechanism thus finds ESO
matches based on the text entered by the modeller as
label for a modelling (i.e., BPMN) element. There
are many ways to compare strings depending on
whether the string is seen as an exact sequence of
letters, an erroneous sequence of letters or a
substring of a set of words. Currently we use the
Jaro-Winkler distance (Winkler, 1990) to calculate
the edit distance between the given BPMN element
label, and the label of each concept in the enterprise-
specific ontology. The Jaro-Winkler distance was
chosen because this hybrid technique takes into
account that the text entered by the modeller can
contain spelling errors, and additionally favours
matches between strings with longer common
prefixes (i.e., a substring test, which is very useful in
our context because matching is executed each time
a character is added to the label).
Jaro-Winkler distances are between 0 (no
similarity) and 1 (exact match), and are thus
immediately useable as a relevance score. For some
modelling constructs, only a part of the label is
matched. For example, tasks’ names in process
models are often specified as a <verb> <noun>
combination (Dumas, La Rosa, Mendling and
Reijers, 2013), and our method only seeks to find
matching for the <noun> part of the label, as only
this part can correspond to a concept in the ontology.
In general, for the following BPMN constructs the
full label is being matched: pool, lane, message flow
and data object. Conversely, for the following
BPMN constructs, only partial matching is
performed: task, sub-process, event and conditional
gateway. The labels of these constructs are analysed
using standard natural language parsing techniques,
which allow identifying the nouns within the labels.
3.2 Synonym Matching Mechanism
The synonym matching mechanism aims to detect
synonyms of the given BPMN element label (or part
of it) in the ESO. To realize this, we use WordNet
(Miller, 1995), an online lexical database that
organises English nouns, verbs, adjectives and
adverbs into sets of synonyms (so-called synsets). It
is thus ideal to find synonyms. For each synonym of
the BPMN element label, the previously described
string matching algorithm is performed on all ESO
concepts, thereby generating a relevance score
between 0 (no match) and 1 (exact synonym match).
Synonym matching allows the modeller to find
concepts that are semantically similar, but for which
he used a different label. For example, while the
modeller enters the label “customer”, the concept
“client” might be available in the enterprise-specific
ontology, and should be given as a suggestion.
Synonyms are semantically equivalent words, but
their usage depends on a certain domain. Therefore,
this mechanism may generate false positives, in case
synonyms in a different context/domain are matched.
As we are operating within one particular domain,
namely that of the enterprise, the chance of false
positives is low.
3.3 Construct Matching Mechanism
The construct matching algorithm finds ESO
suggestions based on the type of modelling
constructs the modeller choses. To do so, the
algorithm exploits the grounding of (i.e., mappings
between) ESO concepts and a core ontology on one
hand, and mappings between the constructs of the
modelling language (i.e., BPMN) meta-model and
the (same) core ontology on the other hand. The core
ontology thus acts as an intermediary between the
ESO and (the meta-model of) the chosen modelling
language. The core ontology that we used in this
paper is the Unified Foundational Ontology (UFO),
for three reasons: 1/ the benefits of grounding
domain ontologies in UFO are well motivated
(Guizzardi, Falbo and Guizzardi, 2008), and several
such UFO-grounded domain ontologies are available,
e.g., (Barcellos, Falbo and Moro, 2010) 2/ UFO is
specifically developed for the ontological analysis of
modelling languages, and 3/ an analysis of BPMN
using UFO is available in literature (Guizzardi and
Wagner, 2011), and can thus be re-used. No claim is
made here that UFO alone is sufficient to find the
match. However, using UFO’s predefined
categories, according to which ESO concepts and
BPMN constructs are classified, allows to guide
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
240
construct-based matching and narrow down the
relevant suggestions. UFO has different layers of
which we here only use UFO-U (represented in
Figure 2), as this is sufficient for finding construct-
based matches between BPMN and ESO. Our
demonstration supports this claim; however, it can
be further investigated if using (full) UFO is
beneficial to refine our algorithm. Additionally,
other modelling languages might require more
specialized UFO concepts, in which case (full) UFO
will be needed. The top level element in UFO-U is a
Universal. It represents a classifier that classifies at
any moment of time a set of real world individuals
and can be of four kinds: Event type, Quality
universal, Relator universal and Object type.
Grounding the enterprise ontology in the UFO
ontology is either done manually, as shown in our
demonstration in section 5, or given when an
existing domain ontology that was previously
grounded, is re-used as initial enterprise ontology.
This effort is thus domain- or even enterprise-
specific. On the other hand, the mapping between
the modelling language’s meta-model and the UFO
ontology is generally applicable, and several
mappings are available in literature, e.g., UML
(Guizzardi and Wagner, 2008), BPMN (Guizzardi
and Wagner, 2011).
We can thus re-use the mapping provided by
(Guizzardi and Wagner, 2011) for UFO - BPMN.
However, in (Guizzardi and Wagner, 2011), the aim
was to perform an in depth ontological analysis of
BPMN, thus providing the mapping between UFO
and BPMN. As we limited the used core ontology to
UFO-U, which is sufficient to find matches between
BPMN and the ESO, we need to adjust the mappings
provided by (Guizzardi and Wagner, 2011). Once
Figure 3: UFO-U.
Figure 4: BPMN meta-model.
SupportingProcessModelDevelopmentwithEnterprise-SpecificOntologies
241
Table 1: Mappings between BPMN constructs and UFO-U.
BPMN construct UFO-U BPMN construct UFO-U
Pool Object type End event Event Type
Lane Object type Event noun Base type, Mixin type,
Relator universal
Task Noun Relator universals or
quality universal
Condition Exclusive
Gateway
Quality universal
Noun Sub Process Relator universals or
quality universal
Data object Relator universal,, Base
type
Start event Event Type Message flow label Relator universal
Intermediate event Event Type
given, this adjusted mapping is re-usable by any
subsequent BPMN modelling efforts. The BPMN
meta-model that is used in this paper is shown in
Figure 4 and is based on the original OMG BPMN
standard.
Table 1 represents the mappings between the
constructs of the BPMN meta-model and UFO-U.
Limiting the match to UFO-U concepts means that
for instance both the Pool and Lane constructs are
mapped to UFO Object types whereas in (Guizzardi
and Wagner, 2011) they are mapped to Agents,
which is a special type of UFO-U Object type. In
(Guizzardi and Wagner, 2011) a task is mapped to
an Action Event Types, which is a special kind of
Event. In this paper we extend this mapping based
on the observation that a lot of well-known BPMN
textbooks such as (Dumas, La Rosa, Mendling and
Reijers, 2013) suggests BPMN modellers to follow
the pattern “verb noun” when specifying the name of
task. The noun that is used corresponds to a UFO-U
relator or quality universal. The mappings of a
subprocess are identical to those of the Task
construct.
The different types of BPMN Events are all
mapped to UFO-U Event Type. It is important to
notice that in most cases the ESO will not contain
UFO-U events, as ESOs most often focus on the
static part of the vocabulary of the enterprise, i.e.,
similar to the view on domain specific ontologies in
(Lin et al., 2006). However, for every BPMN model,
we can search for UFO-U objects that participate in
these events. Gateways are not mapped directly to
UFO-U concepts but for an exclusive data-based
gateway we can expect that the condition that is
evaluated will use properties of UFO Object Types
or UFO Relator Types which in UFO-U correspond
to qualities. Data objects and message flow objects
are most likely to be relators, such as contract, or
base types such as technical documentation of some
software. The association and sequence flow
connectivity objects of the BPMN meta-model are
not mapped to UFO-U, but they will be used
extensively in the construct neighbourhood-based
mechanism (see section 3.4).
To formalize these mappings, we first
transformed both UFO-U and the BPMN meta-
model into OWL ontologies
4
, using the
transformations provided by OMG’s Ontology
Definition Meta-model (OMG, 2006), and
subsequently formalized the mappings between the
constructs of these ontologies using a third OWL
ontology
5
. Using OWL for both representing the
ontologies and the mappings has some advantages: 1)
OWL supports different approaches for specifying
the mapping (OWL equivalent classes, SWRL rules),
2) OWL supports reasoning which will be used to
generate the list of suggestions based on the
provided mappings and 3) OWL uses URIs to
identify the concepts which will be used during the
annotations of the models.
The actual mappings in the OWL mapping
ontology are specified using the OWL equivalent
class construct. The mapping ontology consists of
the UFO-U to BPMN mappings and the UFO-U to
ESO mappings, both of which are specified
separately and subsequently imported. Next, a
reasoner is used to create an inferred class hierarchy
of ESO concepts that are matched to a given BPMN
construct. For instance, when a BPMN pool is
created the inferred class hierarchy is used to
determine all subclasses of BPMN pool (i.e., UFO-U
concepts). Subsequently, all ESO concepts related to
these subclasses (using the UFO-U to ESO mapping)
will receive the relevance score of 1.0, while all the
other ESO concepts will be considered irrelevant
and will receive a relevance score of 0. This
mechanism has been implemented using the OWL
Java API and is available from the GitHub
repository.
   
4
Download from: https://github.ugent.be/MIS/Ontology
BasedSuggestionAlgorithms
5
Download from: https://github.ugent.be/MIS/ Ontology
BasedSuggestionAlgorithms
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
242
3.4 Neighbourhood-based Mechanism
Before the element neighbourhood-based
mechanism is described, we first explain how
ontology annotations are stored, as existing
annotations are used by this mechanism. In order not
to jeopardize the re-usability of our suggestion
generation mechanism among different tools, we
choose not to extend the BPMN meta-model with
annotation elements. Instead, we opted to create, for
every BPMN model, an OWL ontology that contains
instantiations of the BPMN OWL classes. When a
modeller decides to annotate a newly created
element with an ESO concept, a URI of the
corresponding OWL ESO concept is added to the
OWL BPMN individual using the OWL annotation
mechanism. In this way, it is always possible to refer
from the BPMN model elements to the reference
ESO concepts.
With the annotation mechanism explained, we
now detail the element neighbourhood-based
mechanism, which calculates relevance scores for
ESO concepts based on the location of the BPMN
element that is added to the model, the type of
construct that is added, and the relationships
between the ESO concepts. The neighbourhood of a
BPMN element is determined by the connectivity
objects (i.e. sequence flow, message flow,
association), and the lanes recursive relationship of
the BPMN meta-model. In other words, for every
element we can determine which pool or lane it is a
part of, and which other element(s) is/are connected
to this element using either a sequence, message
flow or association. Next, the relationships (which
are specified in terms of the UFO-U relationships
through the ESO-UFO-U mappings) between the
ESO concepts are exploited. According to
(Guizzardi and Wagner, 2008) there are two types of
relations: formal and material. Formal relation
holds between entities directly without any further
intervening individuals. Material relation has
material structure on its own. Entities related by this
type of relation are mediated by individuals called
relators. In section 4 we will demonstrate how the
UFO-U relators can be used to specify the material
relations between the ESO concepts.
Finally, using both the relative position of the
new element and the material relations between the
ESO concepts, the element neighbourhood-based
mechanism can now derive relevance scores for
ESO concepts in relation to some BPMN model
construct (for examples, see section 4):
1. To create a pool construct when another pool
already exists, the suggestions (relevance score 1)
are UFO-U object types that are related by a
material relationship with the ESO concept with
which the existing pool(s) was/were annotated
(i.e., the ontology annotations of the pool(s)).
2. To create a lane construct within a pool, the
suggestions (relevance score 1) are UFO-U
object types that are related by a material
relationship with the ontology annotation of the
pool(s)
3. To create a message construct that results in
transmitting a message between a task or event
of a pool and another pool, the suggestions
(relevance score 1) are UFO-U relators
mediating material relations connecting objects
that annotate respectively the noun of the task
and the ontology annotation of the pool.
4. To create a conditional gateway, there are two
ways to derive suggestions (both receive
relevance score 1):
ESO concepts annotated by the task label
preceding the gateway. This can work very
well for tasks that are performing evaluation
or calculation, after which the gateway is
used to make a decision based on the results.
In this case, the condition on the gateway will
use the same concept as used in the task.
This concept is most likely to be a quality,
especially if the task at hand is performing
calculations. Nevertheless, it can also be a
relator, such as for example verifying if the
contract is ok or not.
Qualities associated with the UFO-U object
type annotation of the pool where the
gateway is located. Or UFO-U qualities
associated with UFO-U object types
participating in material relations with Object
type annotation of the pool where the
gateway is located.
5. For creation of a task construct, the suggested
concepts are most likely to be related through
material relations to the pool where the task is
located. The suggestions can be either object
types or relators mediating those material
relations.
The implementation of this mechanism is more
complex because the algorithm needs to not only
know the location (relative to other elements) of the
new element but also the annotations of the
modelling elements to which the modeller plans to
connect the new element. As mentioned in the
beginning of this section, an OWL ontology file is
created for every model; it contains individuals that
instantiate the BPMN OWL classes. This file, in
particular the instantiations of connectivity concepts,
SupportingProcessModelDevelopmentwithEnterprise-SpecificOntologies
243
is used to determine the position of the newly
created element. Next for selected connected
element (i.e., those satisfying one of the previous 5
rules) the algorithms checks whether it contains an
ESO ontological annotation. In case the surrounding
element is annotated, the algorithm uses a SPARQL
query to identify to which ESO concepts the
ontological annotation is related. The ESO concepts
that are returned by this query consequently receive
a weight of 1.
4 DEMONSTRATION
This section illustrates the method explained in the
previous sections by means of a lab demonstration in
which the modeller constructs a process model in the
Table 2: Mappings between ESO concepts and UFO.
ESO concept UFO-U ESO concept UFO-U
AddedValue Quality_Universal Liability Relator Universal
Mediates Customer and
mediates Branch
Adminstrative Role_Type Loan Relator Universal
Mediates some X and
mediates some X
Asset Mixin Type MortgageLoan Relator Universal
Mediates some X and
mediates some X
Branch Base_Type Payment Relator Universal
Mediates some X and
mediates some X
Channel Relator Universal
Mediates some
Customer and
mediates some Bank
Person Base_Type
Collection Quality_Universal Product Mixin_Type
Commercial Role_Type ProductRateApplication Quality_Universal
Company Base_Type ProductRateApplication
Fixed
Quality_Universal
Corporative Base_Type ProductRateApplication
Fixed
Quality_Universal
Currency Base_Type ProductRateApplication
Variable
Quality_Universal
CurrentMortgageLoan Relator Universal
Mediates some X and
mediates some X
Quota Quality_Universal
Customer Mixin_Type SME Base_Type
Department Base_Type SOHO Base_Type
Employee Role_Type SavingsAccount Base_Type
FutureMortgageLoan Relator Universal
Mediates some X and
mediates some X
Service Mixin_Type
Individuals Base_Type ServiceContractBy
Customer in Chanel
Mixin_Type
InvestmentAccount Base_Type Staff Role_Type
InvestmentFund Base_Type User Mixin_Type
Invoice Relator Universal
Mediates some X and
mediates some X
vBanking Base_Type
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
244
financial domain using an existing financial domain
ontology
6
as enterprise-specific ontology. This
ontology contains static concepts related to finance,
such as Branch, Customer, Loan, Insurance, etc.,
which can be used as a reference for models
constructed in the financial domain. As a first step,
the ESO needs to be mapped to the UFO-U core
ontology. This classification is part of the ontology
engineering cycle of our method and is performed
during the ontology setup phase. Table 2 represents
the mapping, which is formalized in OWL and can
be downloaded from our repository. Important to
notice is that in our mapping, when an ESO concept
is classified as a Relator we also link this Relator to
the Object Types it connects. For example, a relator
Loan is relating Branch and Customer entities. The
ESO also contains OWL data properties, which are
all considered to be quality universals because they
depend on the universals they are related to. For
example, Expiration Date of a Product is a quality
universal because it depends on the universal
Product.
Once the ESO is grounded in the core ontology
(as mentioned, this only needs to be done once, and
can subsequently be re-used for any model created
within the enterprise), the modeller can start creating
the process model. He selects a construct to be
added, places it on the canvas and starts typing the
construct’s desired name. As he selects the construct,
and as he is typing, the mechanisms described in the
previous section derive suggestions from the ESO
and present them to the modeller. If an ESO concept
in the suggestion list appropriately corresponds to
the intention of the modeller for this particular
BPMN construct, he selects this concept, and the
BPMN construct is (automatically) annotated with
the chosen ESO concept.
The process model to be created in our lab
demonstration represents the loan application
assessment process in a bank, and is taken from
(Dumas, La Rosa, Mendling and Reijers, 2013). By
using an existing specification, we avoid bias
towards our method. The process starts when the
loan officer receives a loan application from one of
the bank’s customers. This loan application is
approved if it passes two checks: the first check is
the applicant’s loan risk assessment, which is done
automatically by the system after a credit history
check of the customer is performed by a financial
officer. The second check is a property appraisal
check performed by the property appraiser. After
both checks are completed, the loan officer assesses
   
6
http://dip.semanticweb.org/documents/D10.2eBankingCaseStudy
DesignandSpecificationofApplicationfinal.pdf
the customer’s eligibility. If the customer is found to
be not eligible, the application is rejected. Otherwise,
the loan officer starts preparing the acceptance pack.
He also checks whether the applicant requested a
home insurance quote. If he did, both the acceptance
pack and the home insurance quote are sent to the
applicant. If the insurance was not requested, only
the acceptance pack is sent. The process finally
continues with the verification of the repayment
agreement
Figure 5 represents the BPMN model of the loan
application process. Constructs that are surrounded
by a thick red square are annotated with ESO
concepts. At this stage, a full visual BPMN
modelling tool that implements the suggestion
generation algorithms is still under development.
However, the suggestion algorithm itself, including
all four suggestion mechanisms, is implemented as a
stand-alone engine, and available from the
repository (i.e., the BPMNSuggestionEngine class).
In the remainder of this section the suggestion
generation algorithms described in section 3 is
demonstrated for the different illustrative scenarios.
Adding Branch Pool: the modeller selects the
pool construct to be created. Based on the construct
matching mechanism all ESO concepts
corresponding to UFO-U object type are given a
relevance score of 1 for this matching mechanism.
Among those concepts the modeller can find Branch
which is classified as UFO-U base type. If there is
already a “Customer” pool, based on the construct
neighbourhood matching mechanism, this case
corresponds with rule 1. As the already existing pool
is the Customer pool, the mechanism looks for ESO
concepts related to the Customer concept through
material relationships. There is only one concept
satisfying this requirement: Branch. As a result, the
Branch concept is listed in the beginning of the
suggestion list, as it scored for both the construct
and neighbourhood matching mechanisms (and no
other concept scored equal or higher). Note that in
this scenario, string and synonym matching cannot
contribute to the overall relevance score yet, as the
modeller did not (yet) type any label.
Adding Message flow “loan Application”:
According to the construct matching mechanism,
message flow corresponds to the relator universal.
Therefore, all ESO concepts corresponding to the
UFO-U relator universal will be selected. Those
concepts are: Channel, loan, mortgage loan, current
mortgage loan, future mortgage loan, invoice,
liability, payment. For the element neighbourhood-
based matching technique this situation resolves
under rule 2.
SupportingProcessModelDevelopmentwithEnterprise-SpecificOntologies
245
In our enterprise-specific ontology, all relator
universals are mediating the same two concepts
Branch and Customer. Therefore, the results
delivered by this suggestion generation technique
are the same as the results delivered by construct
matching. In this case, the previously mentioned
suggestions all have equal overall score, and can
thus not be prioritized. We therefore present them
alphabetically. The modeller may select a concept
from the list (i.e., “loan”), or, in case the list is too
long, start typing any desired label (e.g., “loan” or
“credit”). This triggers the string- and synonym-
based matching mechanisms, both of which
prioritize the concept Loan, which consequently
appears on top of the suggestion list, and is selected
by the modeller to annotate the loan application
message flow.
Adding Reject Application Task: The modeller
selects the BPMN task construct, and subsequently
the construct matching mechanism assigns a high
relevance score to all ESO concepts that correspond
to UFO-U quality and relator universals as
suggestions for the task noun. A list of relator
universals is mentioned in the previous example; a
list of quality universals is very exhaustive and is
thus not mentioned here. The second mechanism,
element neighbourhood based matching, applies rule
3: the task at hand is located in the Branch pool, so
this matching mechanism suggests all the ESO
concepts corresponding to UFO-U relator universals
related to Branch concept in the ESO, and UFO-U
object types mediated by those relators (all with
relevance score 1). The Loan concept is a relator
universal, and therefore received relevance score of
both matching mechanisms; it therefore appears on
the top of the suggestions list.
Adding “Home Insurance Quote is Requested”
Gateway: In the last scenario, the modeller draws an
inclusive decision gateway on the canvas. Based on
construct matching mechanism, all quality
universals will be assigned a priority score of 1. The
element neighbourhood-based mechanism classifies
this situation under rule 4, which suggests ESO
concepts that were used to annotate a task construct
preceding the gateway. In this case it is the “check if
home insurance quote is requested” task, which is
annotated with the Home Insurance concept. This
concept thus receives relevance score 1 for the
gateway, and is prioritized in the suggestion list. It
perfectly matches our needs.
To start the discussion, we note that the scenarios
elaborated here were chosen to illustrate the more
complicated cases. As a result, string- and synonym-
based matching mechanism are underrepresented.
Evidently, when no or few BPMN elements are
already on the canvas, neighbourhood-based
matching will be unable to sufficiently differentiate
between potential suggestions (as in scenario 2), and
string- and synonym-matching will become
important. Equally, when the modeller has a certain
label already in mind, string- and synonym matching
will dominate the suggestion list, as the modeller is
typing the label he had in mind. Having made this
comment, we note that in general, it was possible to
derive suggestions based on the construction and
element neighbourhood matching mechanisms for
all model constructs for which the related concepts
existed in the ontology. In fact, as can be seen in the
scenarios, construct and neighbourhood based
matching complement each other well. The majority
of the concepts required by the model, but missing
from the ontology were also correctly classified
under the assumptions of neighbourhood-based
matching mechanism, and would have been assigned
Figure 5: BPMN model describing loan application.
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
246
a high relevance score if they would have been
present in the ontology. However, there was one
case where the neighbourhood-based matching
mechanism was not very accurate. While creating
the last message flow “Home insurance quote”,
based on the second assumption of the
neighbourhood-based matching mechanism and the
construct matching mechanism, relator universals
must be suggested. But in reality, it was annotated
with a quality universal HomeInsurance, instead of a
relator. Further fine-tuning of the suggestion
generation mechanism should avoid this type of
mismatches.
The lab demonstration was used here to
demonstrate viability of our method, to detail the
different steps and provide a concrete case. It shows
that the suggestion generation algorithm indeed
provides useful suggestions to the modeller, and
allows (automatic) annotations of the model, thereby
semantically grounding them and facilitating model
integration. It needs to be mentioned that the lab
demonstration was done using a single modeller, and
that therefore the aforementioned positive
indications of using our method cannot be
statistically proven. We are currently performing a
more elaborate empirical validation, where a group
of 30 test users is divided in three different groups:
one group is given an ESO and our method, the
second group is only given the ESO but without
support of our method, and the last group is not
given an ESO and thus needs to model without any
ontology or method support. The experiment is
specifically designed to show the impact of our
model on modelling efficiency, consistency in the
use of terminology, and the semantic grounding of
the resulting models.
5 CONCLUSIONS AND FUTURE
WORK
In this paper, we detail a suggestion generation
algorithm for BPMN modelling based on an
enterprise-specific ontology (ESO), and showed how
using our modelling approach allows to
automatically annotate the BPMN models with
ontology concepts. This work is part of a larger
ongoing project, which aims to explore the
symbiotic relationship between ontology modelling
on one hand, and business modelling on the other
hand. The suggestion generation algorithm utilizes
four matching mechanisms to derive suggestions
from an ESO. Two of them, the string and synonym
matching mechanisms, are based on the label of the
newly created BPMN element, which is
systematically compared with concepts in the ESO.
The other two, namely construct matching and
neighbourhood-based matching, depend on the type
of the BPMN construct and the position (relative to
other modelling elements) where it is added. A core
ontology, in which the ESO is grounded and with
which the modelling language (BPMN) was
analysed, proves invaluable for these techniques.
Every matching mechanism assigns a relevance
score to every ESO concept; the final relevance
score is calculated based on the weighted average of
scores assigned by every mechanism. This final
relevance score forms the basis for prioritizing ESO
concepts in the suggestion list, which is offered to
the modeller as he is modelling.
Offering the modeller with ESO-based
suggestions aids the modeller, facilitates the
modelling process, promoted cross-model re-use of
ESO concepts, allows model annotation and
ultimately, facilitates model integration.
Future work will follow different directions. First,
we are currently performing further empirical
validation of the benefits of our method. Second,
suggestions towards the ontology (e.g., missing
concepts) based on the modelling process, and
subsequent community-based ontology evolution,
needs to be explored. Finally, we also plan to apply
the meta-method for other modelling languages (i.e.
i*, KAOS), and using other core ontologies.
REFERENCES
Barcellos, M. P. Falbo, R., Dal Moro, R., 2010. A well-
founded software measurement ontology. 6
th
International Conference on Formal Ontology in
Information Systems (FOIS 2010), (vol. 209, pp. 213-
226).
Becker, J., Rosemann, M., Uthmann, C. 2000. Guidelines
of business process modeling. In W. Aalst, J. Desel, &
A. Oberweis (Eds.), Business Process Management
SE - 3 (Vol. 1806, pp. 30–49). Springer Berlin
Heidelberg.
Bera, P., Burton-Jones, A., & Wand, Y., 2011. Guidelines
for designing visual ontologies to support knowledge
identification. Mis Quarterly, 35(4), pp. 883–908.
Euzenat, J., Shvaiko, P., 2013. Ontology Matching. Berlin,
Heidelberg: Springer Berlin Heidelberg.
Fox, M. S., 1992. The TOVE Project: Towards a common-
sense model of the enterprise. (F. Belli & F. J.
Radermacher, Eds.) Industrial and Engineering
Applications of Artificial Intelligence and Expert
Systems. Berlin, Germany: Springer-Verlag.
Fox, M., Gruninger, M., 1997. On ontologies and
SupportingProcessModelDevelopmentwithEnterprise-SpecificOntologies
247
enterprise modelling. Enterprise Engineering and
Integration. pp 190-200. Springer Berlin Heidelberg.
Francescomarino, C., Tonella, P., 2009. Supporting
ontology-based semantic annotation of business
processes with automated suggestions. In T. Halpin, J.
Krogstie, S. Nurcan, E. Proper, R. Schmidt, P. Soffer,
& R. Ukor (Eds.), Enterprise, Business-Process and
Information Systems Modeling SE - 18 (Vol. 29, pp.
211–223). Springer Berlin Heidelberg.
Gailly, F., Laurier, W., Poels, G., 2008. Positioning and
formalizing the REA enterprise ontology. Journal of
Information Systems, 22(2), 219.
Guizzardi, G., Wagner, G., 2008. What’s in a relationship:
An ontological analysis. Lecture Notes in Computer
Science, 5231, pp. 83–97.
Guizzardi, G., Wagner, G., 2011. Can BPMN be used for
making simulation models? In J. Barjis, T. Eldabi, &
A. Gupta (Eds.), Enterprise and Organizational
Modeling and Simulation SE - 8 (Vol. 88, pp. 100–
115). Springer Berlin Heidelberg.
Haller, A., Gaaloul, W., Marmolowski, M., 2008. Towards
an XPDL compliant process ontology. Services - Part
I, 2008. IEEE Congress on.
Klein, M., Bernstein, A., 2001. Searching for services on
the semantic web using process ontologies. In
Proceedings of the International Semantic Web
Working Symposium (pp. 1–16).
Leopold, H., 2013. Natural Language in Business Process
Models. Springer Berlin / Heidelberg.
Liao, Y., Lezoche, M., Loures, E., 2013. Semantic
enrichment of models to assist knowledge
management in a PLM environment. On the Move
toMeaningful Internet Systems, pp. 267–274.
Lin, Y., Krogstie, J., 2010. Semantic annotation of process
models for facilitating process knowledge
management. International Journal of Information
System Modeling and Design, 1(3), pp. 45–67.
Lin, Y., Strasunskas, D., Hakkarainen, S., Krogstie, J.,
Solvberg, A., 2006. Semantic annotation framework to
manage semantic heterogeneity of process models. (E.
Dubois & K. Pohl, Eds.)Advanced Information
Systems Engineering, 4001, pp. 433–446.
Miller, G., 1995. WordNet: a lexical database for English.
Communications of the ACM, 38(11), 39–41.
OMG., 2006. Ontology definition metamodel: OMG
Adopted Specification (ptc/06-10-11). Object
Management Group.
Schlenoff, C., Ivester, R., Knutilla, A., 1998. A robust
process ontology for manufacturing systems. In
Proceedings of 2 nd International Conference on
Engineering Design and Automation.
Sonnenberg, C., Huemer, C., Hofreiter, B., Mayrhofer, D.,
Braccini, A., 2011. The REA-DSL: A domain specific
modeling language for business models advanced
information systems engineering. In H. Mouratidis &
C. Rolland (Eds.), Lecture Notes in Computer Science
6741 (Vol. LNCS 6741, pp. 252–266). Springer Berlin
/ Heidelberg.
Uschold, M., King, M., Moralee, S., Zorgios, Y., 1998.
The enterprise ontology. The Knowledge Engineering
Review: Special Issue on Putting Ontologies to Use,
13(1), pp. 31–89.
Winkler, W., 1990. String comparator metrics and
enhanced decision rules in the Fellegi-Sunter model of
record linkage. In Proceedings of the Section on
Survey Research Methods, American Statistical
Association.
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
248