Evaluating Support for Implementing BPMN 2.0 Elements in Business

Process Management Systems

Carlos Francisco Habekost Dos Santos

1 a

, Marcelo Fantinato

2 b

Jos

e Palazzo Moreira de Oliveira

1 c

and Lucineia Heloisa Thom

1 d

Institute of Informatics, Federal University of Rio Grande do Sul, Porto Alegre, Brazil

School of Arts, Sciences and Humanities, University of S

ao Paulo, S

ao Paulo, Brazil

{cfhsantos, palazzo, lucineia}@inf.ufrgs.br, m.fantinato@usp.br

Keywords:

BPMN, BPMS, Implementation, Notational Elements.

Abstract:

Business Process Model and Notation (BPMN) provides an extensive set of notational elements, such as activ-

ities, events and gateways, which enable the representation of a wide variety of business processes. Business

Process Management Systems (BPMSs) can implement BPMN to allow the execution of business processes.

However, not all BPMN elements are implemented in a BPMS. The purpose of this paper is to present the

results of an evaluation of the BPMN 2.0 elements conducted to identify those whose implementation is sup-

ported in BPMSs. We evaluated four BPMSs, comparing the elements implemented in such BPMSs with

their respective deﬁnitions in the BPMN speciﬁcation, considering the requirements to implement them. As

a result, we found that only 34.18% of the BPMN 2.0 elements are implemented in the investigated BPMSs.

We also identiﬁed that BPMSs implement BPMN elements only partially, adapting their original deﬁnition. In

addition to the results of our evaluation, our contribution is to provide developers with an approach to evalu-

ating the support of BPMN elements in a BPMS. Following the steps proposed in this paper, we identiﬁed the

limitations and devised a method for implementing the remaining BPMN elements not yet implemented.

1 INTRODUCTION

Business Process Management (BPM) provides a set

of techniques for the analysis, implementation, en-

actment and continuous improvement (i.e., evolution)

of business processes in different types of organiza-

tions (Weske, 2014; Dumas et al., 2018). Business

processes

reﬂect the operation of organizations and

allow controlling the development of services and

products that need to be delivered to customers.

Processes can be graphically represented via Busi-

ness Process Model and Notation (BPMN) (OMG,

2013). BPMN has become a standard for process

modeling, disseminated by the Object Management

Group (OMG). BPMN 2.0 is an ISO standard since

2013 (ISO, 2013).

BPMN enables to describe real-world processes,

https://orcid.org/0000-0003-3291-5208

https://orcid.org/0000-0001-6261-1497

https://orcid.org/0000-0002-9166-8801

https://orcid.org/0000-0002-0620-9302

For the sake of simplicity, in this paper, we use only the

term process to refer to the expression business process.

which can be supported by BPMN tools (Cortes-

Cornax et al., 2015). For business practitioners,

BPMN enables internal communication of business

procedures, business-IT alignment and collaboration

among business partners (Salles et al., 2013; Arevalo

et al., 2016; Salles et al., 2018).

A process model in BPMN is called process di-

agram, and two or more processes interacting with

each other can form a collaboration diagram. Col-

laboration diagrams can comprise the following set

of basic notational elements

: activities, events, gate-

ways, groups, annotations, data objects, databases,

pools, lanes, sequence ﬂows, message ﬂows and asso-

ciations).

BPM life cycle phases include process modeling,

conﬁguration, implementation, execution and valida-

tion (Weske, 2014; Dumas et al., 2018). During mod-

eling, we can model a process in BPMN (OMG,

2013). The process modeled in BPMN can then be

implemented, i.e., be further detailed at a granular

level that allows its automated execution. These ac-

For the sake of simplicity, in this paper, we use only the

term element to refer to the expression notational element.

Santos, C., Fantinato, M., Moreira de Oliveira, J. and Thom, L.

Evaluating Support for Implementing BPMN 2.0 Elements in Business Process Management Systems.

DOI: 10.5220/0009356501930200

In Proceedings of the 22nd International Conference on Enterprise Information Systems (ICEIS 2020) - Volume 2, pages 193-200

ISBN: 978-989-758-423-7

193

tivities can be supported by a BPM System (BPMS).

A BPMS aims to coordinate an automated process

model so that all work is done at the right time and by

the right resource (Weske, 2014; Dumas et al., 2018).

Automating a BPMN process model requires pro-

viding the model with implementation details that are

not required when it comes to only describing process

models for business purposes. These implementation

details are required to remove ambiguities and need

to be accurate so that the process model can be inter-

preted by a BPMS (Gassen et al., 2015; Dumas et al.,

2018; Santos et al., 2019). For example, if a process

model contains abstract tasks, one must specify their

attributes, documents and resources.

The lack of detailed information on implementa-

tion aspects in BPMN represents a technical challenge

to transforming a process model into an executable

version. As a result, each BPMS supplier uses its pro-

prietary format to map and transform a process model

into an executable model (Geiger et al., 2013). Conse-

quently, the adherence of current BPMSs, considering

the BPMN elements they implement, to the BPMN

speciﬁcation is unknown.

In this paper, we present the results of an eval-

uation of BPMSs considering their degree of imple-

mentation of BPMN 2.0 elements. Our goal is to

measure and show the completeness of BPMN imple-

mentations in BPMSs. Supported by our approach,

BPMS developers can identify missing elements in

their BPMSs to implement them as well as verify

whether the present elements are implemented fol-

lowing the BPMN 2.0 speciﬁcation so that noncon-

forming elements can be adjusted.

For example, considering the BPMN element

message ﬂow, a BPMS supplier can verify whether

its BPMS implements it. If the element is miss-

ing, the BPMS developer may implement it. On the

other hand, if this element is already implemented,

the BPMS developer can verify the element behav-

ior. If the element implementation does not follow the

BPMN 2.0 speciﬁcation, the developer must adjust it

to comply with the ofﬁcial speciﬁcation.

The protocol for evaluating the BPMN elements

implemented in BPMSs is summarized by these steps:

(i) We chose free BPMSs to be evaluate; (ii) We de-

ﬁned as scope the BPMN elements used for collab-

oration diagrams as deﬁned in the BPMN 2.0 spec-

iﬁcation (ISO, 2013); (iii) We modeled a process in

each selected BPMS verifying whether the elements

were implemented in the BPMSs and for the imple-

mented elements, we veriﬁed whether they followed

the BPMN 2.0 speciﬁcation; (iv) To systematize this

evaluation, we set a score for the implemented ele-

ments of each BPMS according to the level of adher-

ence to the BPMN speciﬁcation.

Our main contribution is to present an evaluation

of whether and how BPMSs implement BPMN ele-

ments, with respect to the BPMN 2.0 speciﬁcation.

Some insights related to this evaluation could be ob-

tained, including: the proﬁle of BPMN elements com-

monly unimplemented in BPMSs, the limitations of

current BPMN implementations in BPMSs and future

research directions focusing on improving BPMSs

and the BPMN speciﬁcation.

The remainder of this paper is structured as fol-

lows: Section 2 presents the related work; Section 3

details the protocol deﬁned to select and evaluate the

BPMSs; Section 4 depicts the results obtained; and

Section 6 summarizes the approach, discusses its lim-

itations and concludes the paper.

2 RELATED WORK

In this section, we analyzed studies related somehow

to model transformation targeting at execution or sim-

ulation models.

Table 1 summarizes the related work discussed in

this section. In summary, we found three types of ap-

proaches: (a) BPMS limitations, i.e., studies whose

goal is to evaluate overall BPMS limitations; (b)

model transformation limitations, i.e., studies whose

goal is to evaluate limitations on the model trans-

formation; (c) model transformation proposal, i.e.,

studies whose goal is to propose new approaches for

model transformation.

Table 1: Comparison of Related Work [(a) BPMS Limi-

tations; (B) Model Transformation Limitations; (C) Model

Transformation Proposal].

Authors & year (a) (b) (c)

orger (2012) X

Cetinkaya et al. (2012) X

Peralta et al. (2014) X

Geiger et al. (2013) X

Bocciarelli et al. (2014) X

Kluza et al. (2015) X

Meidan et al. (2017) X

orger (2012) evaluated the BPMN 2.0 speciﬁcation

and found many behavioral issues that the speciﬁca-

tion leaves open. The issues described include the

BPM life cycle concept that does not characterize

the mechanism of interruption and compensation for

transactions; the expression “evaluate” is not clear,

because it is not deﬁned when a condition deﬁned in

any part of the process model can be evaluated; a gen-

eral notion of state is missing and hence the deﬁni-

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

194

tion of data dependent conditions is only poorly sup-

ported. As a consequence of these issues, according

to these authors, the BPMS suppliers typically imple-

ment only subsets of the so-called standard and still

are often only partially compatible with each other.

Meidan et al. (2017) conducted a survey to evalu-

ate BPMSs and highlight each phase of the BPM life

cycle fully supported by them. The survey combined

Systematic Literature Review (SLR) and quality mod-

els. SLR was used to select the BPMSs. The follow-

ing BPMSs were selected: Activiti, Bonita, jBPM,

ProcessMaker, uEngine and Camunda. To evaluate

the selected BPMSs, the authors considered as quality

models criteria related to modeling, design, deploy-

ment, execution & control and analysis. As the main

result, these authors presented an evaluation present-

ing the BPMS closest to the aims of the BPM life cy-

cle when compared to other BPMSs.

Geiger et al. (2013) studied the issues in BPMN

serialization that arise due to the complexity and in-

consistency of the BPMN speciﬁcation. Serialization

is to translate data structures or objects into a format

that can be stored and reconstructed later. The authors

considered the Web Services Business Process Execu-

tion Language (WS-BPEL) as a model serialization.

WS-BPEL is a serialization format for an executable

service-based process model. The BPMN speciﬁca-

tion provides a mapping from the process model in

BPMN to WS-BPEL. However, these authors depict

that the BPMN speciﬁcation does not provide the cor-

rect serialization format, making it difﬁcult to turn

a process model executable. The approach provides

two contributions: a list of relevant constraints in

the BPMN speciﬁcation to turn a process model ex-

ecutable and the respective serialization.

Cetinkaya et al. (2012) proposed an approach to

transform a process model into a simulation model.

These authors developed a framework that minimizes

the gap between these models. The framework was

developed based on Discrete Event System Speciﬁca-

tion (DEVS), which is a mathematical formalism used

to represent systems. Models represented in DEVS

are called atomic models. These atomic models are

deﬁned with the following information: the set of in-

put values, the set of output values, the set of states,

the internal transition function, the external transition

function, the output function and the time advance

function. For each element in BPMN, there is a corre-

sponding element target in DEVS. As the main result

of this transformation, one can obtain a model that can

be executable at the simulation level.

Peralta et al. (2014) proposed metrics to analyze

and measure a process model, aiming to improve and

facilitate its implementation in the cloud. These au-

thors considered as metrics: number of activities, the

total number of precedence dependencies between ac-

tivities, connectivity level between activities, num-

ber of split nodes of parallel, number of XOR-split,

number of OR-split, number of AND-join, number

of XOR-join and number of OR-join. These met-

rics were applied in the process model, in a semi-

automated way, and resulted in a description of its

needs to implement a process model in the cloud.

Bocciarelli et al. (2014) proposed an approach to

transform process models to the simulation level. To

this end, these authors developed a Java-based tool

that, taking as input a BPMN 2.0 process model, ob-

tain, as a result, a simulation model, called eBPMN.

The eBPMN core includes a set of BPMN elements,

adapted to the simulation model. According to these

authors, this model allows the process model to be ex-

ecuted with details that are needed to simulate it.

Kluza et al. (2015) focused on the semantic of the

process model. To achieve this, the authors applied

ontology concepts on free BPMSs (Activi, jBPM and

Camunda) to increase the semantic representation.

According to these authors, applying semantic mod-

eling allows disambiguation of data description and

control of its integrity. As a result, this approach al-

lows data transformation, during translating a process

model into an execution model.

The approaches described above have different

types of goals, including: evaluating the BPMN

2.0 speciﬁcation, evaluating BPMSs in terms of

their BPMN implementation, identifying limitations

to transforming a process model into an executable

model, or proposing an approach to transforming a

process model into an executable model.

As for the evaluation of BPMSs regarding the

BPMN elements implemented by them, the works

found do not address the evaluation of the implemen-

tation of the elements per se. Approaches focused on

the evaluation of the BPMN speciﬁcation also do not

verify the elements. Works that propose new model

transformation approaches can lead to an even greater

difﬁculty in understanding process models, adding

more concepts to be understood beyond BPMN. On

the other hand, the BPMN element evaluation pro-

posed herein can be seen as a way to provide a set

of elements that can be useful to users if their goal is

to automate their process models.

3 RESEARCH PROTOCOL

This section presents the research protocol of our

evaluation of the implementation of BPMN elements

in BPMSs. Our goal is to present the steps we fol-

Evaluating Support for Implementing BPMN 2.0 Elements in Business Process Management Systems

195

Figure 1: Phases of the Evaluation of BPMN Elements in BPMSs.

lowed as an approach so that they can be followed

by those looking to evaluate other BPMN elements or

BPMSs in the future.

Brieﬂy, our approach comprised three phases:

1. Selecting the set of BPMSs to be evaluated.

2. Selecting the BPMN elements for evaluation.

3. Evaluating the implementation of BPMN ele-

ments in the selected BPMSs (verifying whether

the BPMN elements are implemented in the

BPMSs and verifying whether the implemented

BPMN elements are correctly implemented).

4. Scoring the evaluated BPMSs according to which

(and how) BPMN elements are implemented.

Figure 1 depicts the conducted phases. First, we se-

lected a set of BPMSs to be evaluated (cf. “Phase 1”),

considering only free BPMSs. Then, we selected a

set of BPMN elements (cf. “Phase 2”), for which we

considered only elements used for collaboration di-

agrams as deﬁned in the BPMN speciﬁcation (ISO,

2013). Based on the set of selected BPMN elements,

we evaluated one by one for each selected BPMS to

verify: whether each BPMN element is implemented

in the BPMS (cf. “Phase 3.1”); and whether the im-

plemented BPMN elements are implemented accord-

ing to the BPMN speciﬁcation (cf. “Phase 3.2”). Re-

sults are counted with a score that deﬁnes the level of

adherence of each BPMS to the BPMN speciﬁcation

(cf. “Phase 4”). Finally, we analyzed the results.

3.1 Selecting the BPMSs

The ﬁrst phase of the evaluation is the selection of

BPMSs. To this end, we considered the following

criteria: (i) the BPMS must be listed in the BPMN

ofﬁcial website

, present in the “Implementers” list;

(ii) the BPMS must be free software; and (iii) the

http://bpmn.org

BPMS must be able to implement and execute pro-

cesses modeled in BPMN 2.0.

We used only free BPMSs following approaches

found in related work (Meidan et al., 2017; Corra-

dini et al., 2018; Kluza et al., 2015). The use of free

BPMSs improves the reproducibility of the evaluation

presented in this paper. As for the followed proce-

dure, we veriﬁed the BPMN website for information

about the tool suppliers. If a free version of the BPMS

was available, we downloaded it. This approach can

also be applied for the commercial versions, making

our study applicable for a full set of BPMSs.

In addition, we considered only BPMSs that al-

low implementing and executing processes modeled

in BPMN 2.0. By execution, we mean that the BPMS

must be able to carry out the process, by instantiat-

ing and controlling its performance. We selected the

BPMSs on March 2018.

Finally, Table 2 depicts the four BPMSs selected.

Considering all deﬁned criteria, only four BPMSs jus-

tify a closer evaluation regarding the degree to which

they actually implement the BPMN speciﬁcation.

Table 2: Final Selected BPMSs.

BPMS Supplier Ver-

sion

Release

date

Bonita BPM Bonitasoft 7.6.2 Jan, 2018

Camunda BPM Camunda 7.9.0 May, 2018

jBPM KIE Group 7.7 Mar, 2018

Web Ratio WebRatio 8.8.1 Not found

3.2 Selecting the BPMN Elements

In the second phase, we selected the set of BPMN 2.0

elements to be used in the evaluation. To this end,

we considered only the BPMN elements used for col-

laboration diagrams as deﬁned in the BPMN speciﬁ-

cation. Thus, we did not consider the elements used

speciﬁcally for conversation diagrams and choreog-

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

196

raphy diagrams. Collaboration diagrams are those

more commonly used by BPMN practitioners, which

justiﬁes our decision. The set of elements chosen al-

lows the reader of a process model in BPMN to easily

recognize the types of elements used and understand

the diagram (ISO, 2013). In total, there are 83 ele-

ments, considering both the basic elements and their

extended versions, which are grouped in ﬁve cate-

gories – ﬂow objects, artifacts, data, swimlanes and

connecting objects.

4 EVALUATING THE

IMPLEMENTATION OF BPMN

ELEMENTS IN BPMS

Once the BPMS and BPMN elements to be evaluated

were selected, we proceeded with the third phase of

the research protocol (cf. Section 3) and conducted

the evaluation itself. To this end, we evaluated the

BPMN implementations in two aspects:

1. Verifying whether the BPMN elements are imple-

mented in the BPMSs.

2. Verifying whether the implemented BPMN ele-

ments are correctly implemented.

4.1 Identifying Implemented BPMN

Elements

For this ﬁrst evaluation aspect, we created a process

model on each BPMS under evaluation and tried to

add to it each one of the BPMN elements selected for

evaluation (cf. Section 3.1). Thus, we seek to iden-

tify elements not implemented in BPMSs, i.e., those

that would not be possible to be added to the process

model as they were not available in the BPMS.

To select the elements, ﬁrst, we veriﬁed the editor

of each BPMS. If the BPMN element was available

for use in the BPMS editor, we concluded that the

corresponding BPMS implements it. We also tried

to create extended versions of elements through the

element properties, for example by right-clicking on

the element. This is often available in some tools for

changing the type of some elements, e.g., changing

an abstract task to a script task or a standard event

to a message event. If there was no reference to the

evaluated BPMN element in the BPMS editor or in a

base element properties, we concluded that the corre-

sponding BPMS does not implement it. To control the

conducted evaluation, we used a spreadsheet where

we marked an “X” for each implemented BPMN el-

ement. We organized this spreadsheet by BPMN ele-

ment groups.

4.2 Evaluating Implemented BPMN

Elements

For the second evaluation aspect, we modeled differ-

ent processes by incrementally using all the BPMN

elements implemented in each BPMS, as identiﬁed

according to Section 4.1.

After modeled the process, we followed a proce-

dure for obtaining a quantitative assessment for the

following groups of elements: activities, events, gate-

ways, data, connection objects, swimlanes and arti-

facts. This procedure allowed a systematic evaluation

of the implementation of the BPMN elements and a

homogeneous comparison between the BPMSs.

The procedure for evaluating the implemented

BPMN elements consists of: Element score: we as-

signed 0, 1 or 2 points to each BPMN element, con-

sidering that 0 points means that the BPMS does not

implement the BPMN element, 1 point means that the

BPMN element is partially implemented considering

the BPMN speciﬁcation and 2 points means that the

BPMN element is fully implemented considering the

BPMN speciﬁcation. For example, considering the

part of the deﬁnition of the XOR split-gateway ele-

ment that states “The ﬁrst condition that evaluates to

true determines the sequence the token is sent to and

no more conditions are henceforth evaluated”: if the

BPMS does not implement this part of the deﬁnition,

it receives a maximum of 1 point; but if the BPMS

implements this part of the deﬁnition, it can receive

2 points, depending on whether or not the rest of the

deﬁnition of this element is implemented.

Element group score: for each group of BPMN el-

ements, we calculated (cf. Equation 1) a Group Score

(GS) by summing the individual scores of all BPMN

elements for the group and dividing the result by the

number of elements in the group. In Equation 1, n

represents the number of elements in the group and S

represents the individual score of the element.

GS =

∑

i=1

(1)

Normalized group score: for all element group scores

previously calculated (cf. Equation 1), we calculated

(cf. Equation 2) a Normalized Group Score (GS

norm

)

in the 0-10 range, dividing each group score by the

highest score obtained by all element groups, and then

multiplying the result by 10. In Equation 2, n repre-

sents the number of element groups, GS represents

the group score and max(GS) represents the highest

among the group scores.

Evaluating Support for Implementing BPMN 2.0 Elements in Business Process Management Systems

197

Table 3: Summary of BPMS Evaluation.

BPMN element group Bonita BPM Web Ratio BPM Camunda BPM jBPM

Activities 8.33 5.00 8.89 7.78

Events 7.04 7.41 6.67 6.48

Gateways 7.00 9.00 7.00 7.00

Connection objects 8.00 4.00 6.00 4.00

Artifacts 10.00 7.50 10.00 5.00

Swimlanes 10.00 10.00 10.00 10.00

BPMS

score

8.40 7.15 8.09 6.71

9 9

5.56

7.78 7.78

8.89

2.94 2.94 2.94

3.53

10 10

4.29

7.14 7.14 7.14

10 10 10 10

4.17 4.17

3.33

2.5

4.17 4.17 4.17

6.25

Start Events

Swimlanes

End Events

Gateways

Intermediate Events

Activities

Artifacts

Connection Objects

Percentage implemented

BPMS

Bonita BPM

Camunda BPM

jBPM

WebRatio BPM

Figure 2: Percentage of Implemented BPMN Elements by Element Group and by BPMS.

norm

max(GS)

x10, ∀i ∈ 1..n (2)

BPMS score: for each BPMS, we calculated (cf.

Equation 3) the BPMS

score

by summing all the group

scores for the BPMS and dividing the result by the

number of element groups. In Equation 3, n rep-

resents the number of BPMN element groups and

norm

represents the normalized group score.

BPMS

score

∑

i=1

norm

(3)

Table 3 summarizes the evaluation results. Each row

represents a group score (cf. Equation 1), for the

evaluations of the BPMN elements of the correspond-

ing group. For example, row “Activities” depicts the

scores for this group of BPMN elements, for each

BPMS evaluated. The BPMS

score

(cf. Equation 3)

row depicts the ﬁnal scores obtained by BPMS, con-

sidering the six groups above.

5 RESULT ANALYSIS

Our study found that support for BPMN elements in

BPMSs is limited considering the four BPMSs evalu-

ated. Of the 79 elements evaluated, only 27 are imple-

mented by at least one of the BPMSs (corresponding

to 34.18%). We also identiﬁed that different BPMSs

usually implement the same set of BPMN elements.

For example, in terms of gateways, BPMSs usually

implement AND, XOR and OR gateways while the

event-based gateway is not implemented. One hy-

pothesis is that only the BPMN elements most often

used by practitioners are implemented in BPMSs.

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

198

Bonita BPM Camunda BPM jBPM WebRatio BPM

BPMS

Number of elements

Not

implemented

Partially

implemented

Fully

implemented

Figure 3: Number of Implemented BPMN Elements by BPMS.

Fig. 2 presents more information about the results ob-

tained by each group of BPMN elements. We split

events into three groups to allow a more precise result

analysis: start events, intermediate events and end

events. As a result, one can observe that “start events”

is the element group with the fewest elements im-

plemented (average of 30.88%). In contrast, “swim-

lanes” is the element group with the most elements

implemented (average of 100%) followed by “arti-

facts” (average of 62.5%). One hypothesis for the

higher level of implementation of these two groups is

their possible ease of implementation as both groups

represent simpler elements in terms of behavior rules

according to the BPMN speciﬁcation.

Fig. 3 depicts the results grouped by BPMS, con-

sidering the elements that there are an implementa-

tion in the BPMS. We present an analysis showing

the number of elements fully implemented, partially

implemented and not implemented according to the

BPMN speciﬁcation. From this data, one can con-

clude that Bonita BPM is the BPMS with the largest

number of BPMN elements fully implemented, con-

sidering the BPMN speciﬁcation. From 50 elements,

35 are fully implemented (42.17%), 6 are partially im-

plemented (7.23%) and 9 not implemented (10.84%).

Regarding the scope of implementation, we can

conclude that although the BPMS implements few

elements, not all BPMS perform as deﬁned in the

BPMN speciﬁcation. While Bonita BPM obtained the

highest score (about 8.40), jBPM obtained the low-

est score (6.71). This means that Bonita BPM offers

higher support to BPMN, compared to another tool.

About the evaluated groups, we can observe that

BPMSs in general support swimlanes (all scored 10)

and artifacts (Bonita and Camunda scored 10; We-

bRatio 7.5 and jBPM 5). In opposition, activities,

events and gateways are weakly supported.

6 CONCLUSION

In this paper, we presented an evaluation of the cur-

rent state of the support of the BPMN elements in

available BPMS. To achieve this aim, we focused on

a protocol that allows us to investigate the BPMS.

With the information about the implementation, we

can identify the elements that need to be implemented

and hence investigate how to do it.

According to our evaluation of related works, the

available investigations are focused on verifying re-

source aspects of the BPMS. Our research can help

to determine the limits of support, considering the

BPMS. With this, one can develop strategies to im-

plement the remained unimplemented elements. With

these elements implemented, it is possible to incre-

ment the number of element offers.

Our main contribution is an approach to evaluate

the support of elements on a set of selected BPMS.

Following the steps proposed in this paper, it is possi-

ble to identify limitations on a BPMS in development

and plan a way to implement the remaining elements

that there is no existent implementation. Applying the

protocol, we obtained a set of 27 elements that there

is at least one BPMS that implements it. With the 27

elements, we veriﬁed the implementation, revealing

that there are elements that the implementation does

Evaluating Support for Implementing BPMN 2.0 Elements in Business Process Management Systems

199

not follow the BPMN speciﬁcation.

We identiﬁed that the BPMS evaluated in this re-

search does not cover all BPMN elements. For exam-

ple, this protocol may be applied to Choreographies

and Conversations. About the selection of BPMSs,

considering many BPMSs are a not-free license, this

condition reduced our number of BPMS evaluated.

We cannot assure if these BPMSs provide support to

the elements that there is no implementation.

As future work, we suggest: (i) analyze the lack

of semantics with the elements that there is no im-

plementation; (ii) analyze if the missing elements

will increase the difﬁculty to understand the process

model, (iii) analyze the relation about modeling errors

with the lack of elements and (iv) propose pseudo-

algorithm to the missing elements, aiming to facilitate

the development of these elements.

ACKNOWLEDGEMENTS

Jos

e Palazzo Moreira de Oliveira receive support from

CNPq by grants 301425/2018-3 and 400954/2016-8;

Carlos Francisco Habekost dos Santos is a scholarship

holder from CNPq; this study was ﬁnanced in part by

the CAPES - Brazil - Finance Code 001.

REFERENCES

Arevalo, C., Escalona, M. J., Ramos, I., and Dom

ınguez-

noz, M. (2016). A metamodel to integrate business

processes time perspective in BPMN 2.0. Information

and Software Technology, 77:17–33.

Bocciarelli, P., D’Ambrogio, A., Giglio, A., Paglia, E., and

Gianni, D. (2014). Empowering business process sim-

ulation through automated model transformations. In

Proceedings of the Symposium on Theory of Modeling

& Simulation - DEVS Integrative, DEVS ’14, pages

1–9.

orger, E. (2012). Approaches to modeling business pro-

cesses: A critical analysis of bpmn, workﬂow pat-

terns and yawl. Software and Systems Modeling,

11(3):305–318.

Cetinkaya, D., Verbraeck, A., and Seck, M. D. (2012).

Model transformation from bpmn to devs in the

mdd4ms framework. In Proceedings of the 2012 Sym-

posium on Theory of Modeling and Simulation - DEVS

Integrative M&S Symposium, TMS/DEVS ’12, pages

1–6.

Corradini, F., Ferrari, A., Fornari, F., Gnesi, S., Polini, A.,

Re, B., and Spagnolo, G. O. (2018). A guidelines

framework for understandable bpmn models. Data &

Knowledge Engineering, 113:129–154.

Cortes-Cornax, M., Matei, A., Dupuy-Chessa, S., Rieu, D.,

Mandran, N., and Letier, E. (2015). Using intentional

fragments to bridge the gap between organizational

and intentional levels. Information and Software Tech-

nology, 58:1–19.

Dumas, M., Rosa, M. L., Mendling, J., and Reijers, H. A.

(2018). Fundamentals of Business Process Manage-

ment, Second Edition. Springer.

Gassen, J. B., Mendling, J., Thom, L. H., and Palazzo Mor-

eira de Oliveira, J. (2015). Towards guiding process

modelers depending upon their expertise levels. In Ad-

vanced Information Systems Engineering Workshops

- CAiSE 2015 International Workshops, Stockholm,

Sweden, June 8-9, 2015, Proceedings, pages 69–80.

Geiger, M., Wirtz, G., and der Weberei, A. (2013). Bpmn

2.0 serialization-standard compliance issues and eval-

uation of modeling tools. In International Workshop

on Enterprise Modeling and Information Systems Ar-

chitectures, pages 177–190. Citeseer.

ISO (2013). ISO/IEC 19510:2013: Information technology

– Object Management Group Business Process Model

and Notation, V2.0.2. Standard, International Organi-

zation for Standardization, Geneva.

Kluza, K., Kaczor, K., Nalepa, G. J., and

Sla

nski, M.

(2015). Opportunities for business process semantiza-

tion in open-source process execution environments.

2015 Federated Conference on Computer Science and

Information Systems (FedCSIS), pages 1307–1314.

Meidan, A., Garca-Garca, J., Escalona, M., and Ramos, I.

(2017). A survey on business processes management

suites. Computer Standards & Interfaces, 51(C):71–

86.

OMG, O. M. G. (2013). Business process model and nota-

tion (bpmn) version 2.0. Standard.

Peralta, M., Salgado, C., Baigorria, L., Riesco, D., Mon-

tejano, G., Debnath, N., and Hu, J. (2014). Work-

ﬂow models: Management and quality of process in

the cloud. In International Conference on Informa-

tion Technology: New Generations, pages 91–96.

Salles, G. B. M., Fantinato, M., Barros, V. A., and Albu-

querque, J. P. (2018). Evaluation of the StrAli-BPM

approach: Strategic alignment with BPM using agree-

ments in different levels. International Journal of

Business Information Systems, 27(4):433–465.

Salles, G. B. M., Fantinato, M., Nishijima, M., and Al-

buquerque, J. P. (2013). A contribution to organiza-

tional and operational strategic alignment: Incorpo-

rating business level agreements into business process

modeling. In Proc. of the 10th Int. Conf. on Services

Computing, SCC 2013, pages 17–24. IEEE.

Santos, C. F. H. D., Thom, L. H., Cota,

E., and Fantinato,

M. (2019). Supporting bpmn tool developers through

meta-algorithms. International Journal of Business

Information Systems, 32(4):460–488.

Weske, M. (2014). Business Process Management.

Springer, Berlin Heidelberg, 2 edition.

ICEIS 2020 - 22nd International Conference on Enterprise Information Systems

200