APPLICATION OF SERVICE-ORIENTED COMPUTING

AND MODEL-DRIVEN DEVELOPMENT PARADIGMS TO

BUSINESS PROCESSES

A Systematic Review

Andrea Delgado

Computer Science Institute, Faculty of Engineering, University of the Republica, Montevideo, Uruguay

Francisco Ruiz, Ignacio García-Rodríguez de Guzmán, Mario Piattini

Alarcos Research Group, Technologies and IS Department, University of Castilla-La Mancha, Ciudad Real, Spain

Keywords: Business Process Management (BPM), Service-oriented Computing (SOC), Model-driven Development

(MDD), Systematic Review.

Abstract: To achieve the defined value for their businesses, current organizations need to manage their business

processes in an integrated manner, interconnecting the software systems that support these processes. Over

the last few years, new paradigms have appeared to respond to this and other organizational and software

needs: Business Process Management (BPM) and Service-Oriented Computing (SOC) which are closely

interconnected. Additionally, the Model-Driven Development (MDD) paradigm has been called upon to

play an important role in supporting business process implementation by software services. BPM handles

the management of business processes, including their modelling, deployment, execution, analysis and

improvement. Service-Oriented Computing bases software development on services, which correspond to

business concepts and are created in order to perform business processes. Model-Driven Development

promotes software development based on models which enable, among other things, transformations and the

automatic generation of code for different platforms. With the aim of establishing the bases for research into

the integration of these paradigms to support business process management in organizations, a systematic

review was carried out, focusing on the current state of the literature concerning the application of service-

oriented and model-driven paradigms to business processes.

1 INTRODUCTION

The software integration effort towards realizing

horizontal business processes that give value to

business has been an important issue over the last

few years. Due to the verticality and complexity of

software systems generally implemented in different

technologies, integrating them has been a difficult

task. One challenge is to unite the vision from

business and software areas, to design and

implement business processes in a way that allows

us to react agilely to changes, mainly in two aspects:

business and technologies. New paradigms have

appeared to support this: Business Process

Management (BPM), Service-Oriented Computing

(SOC) and Model-Driven Development (MDD).

BPM refers to the set of activities that organizations

perform to optimize or adapt their business

processes to new organizational needs; BPM

Systems (BPMS) are tools supporting their integral

management, from modelling to execution in a

process engine (BPMI) (Smith et al, 2003). SOC

bases the design of applications on services, which

are software reusable elements through which

providers and consumers interact in a decoupled way

to perform business processes in defined sequences

of invocations to services (orchestration,

choreography). SOA is a realization of SOC (Erl,

2005) (Krafzig et al, 2005) (Papazoglou et al, 2007).

Model-Driven Development (MDD) bases software

development on models, whose elements are

described using metamodels, models and languages

which allow the automatic transformation between

them along with the generation of code for different

Delgado A., Ruiz F., García-Rodríguez de Guzmán I. and Piattini M. (2010).

APPLICATION OF SERVICE-ORIENTED COMPUTING AND MODEL-DRIVEN DEVELOPMENT PARADIGMS TO BUSINESS PROCESSES - A

Systematic Review.

In Proceedings of the 5th International Conference on Software and Data Technologies, pages 51-62

DOI: 10.5220/0003002700510062

 SciTePress

platforms. MDA is a specific realization of MDD

(MDA, 2003) (Mellor et.al, 2003) (Stahl et al., 2006)

In this article, a systematic review of the

application of service-oriented and model-driven

paradigms to business processes is presented, with

the main objective of providing the basis for the

research into business process improvement. The

research path we have defined is that of the

methodological application of service-oriented and

model-driven paradigms to business processes. The

focus is on disciplined, conceptually based and

standardized associated practices, not on a specific

tool or interpretation. This will allow us to maximize

the value of the understanding and application of

these paradigms jointly, to obtain an organization’s

defined business value. The rest of the article is

organized as follows: section 2 briefly describes the

procedure followed in carrying out the systematic

review and its main elements; in section 3, relevant

information extracted from the selected primary

studies is presented, and in section 4, the main

principles in paradigm integration identified in the

analysis of the studies are discussed. Finally, section

5 presents some conclusions.

2 REVIEW PROCEDURE

Following the method proposed by Kitchenham

(Kitchenham, 2004), a systematic review consists of

three stages: planning the review, development of

the review and publication of the results. These

stages have several elements, starting from the

definition of the research question, key words and

research chains, the execution of sources for the

defined chains, and the inclusion and exclusion

criteria to select relevant and primary studies from

which to extract the associated data. The defined

objective is to identify the joint application of SO

and MD paradigms to business processes, which

allows us to assess aspects related to their use.

2.1 Research Question

The research question that guides this work includes

these specific terms in its initial formulation: SOC,

MDD, MDE, BP and BPM. In the update done in

June 2009, it was extended to include new terms like

Service-Oriented Development (SOD). The research

question, then, is:

Which methodological/conceptual applications

of Service-Oriented (Computing or Development)

and Model-Driven (Development or Engineering)

paradigms to business processes and the Business

Process Management (BPM) paradigm have been

carried out?

The main interest is in the methodological and

conceptual levels of application, not in technological

aspects or specific tools for implementation. The

main identified key words for carrying out the

searches are: Service-Oriented Computing and

Development (SOC, SOD), Model-Driven

Development and Engineering (MDD, MDE),

Business Process and Business Process Management

(BP, BPM). The research question, keywords and

search strings were validated by experts.

2.2 Search String and Sources

The criteria for selecting the sources included the

importance of the source as a repository of scientific

articles, possibility of access and web site, resulting

in the following selections: ACM Digital Library

(http://portal.acm.org), IEEE Digital Library (http://

www.computer.org), SCOPUS document db (http://

www.scopus.com), Science@Direct in CS area

(http://www.sciencedirect.com), WileyInterScience

in CS area (http://www.interscience.wiley.com).

The search strings were obtained from the

combination of the defined keywords, mainly with

the Boolean operands “AND” and “OR”. The main

search string was constructed by combining the

defined key words, obtaining the following general

search string, which was adapted for each search

engine of the selected sources:

("service-oriented computing” OR “service-oriented

development”) AND ("model-driven development”

OR "model-driven engineering”) AND ("business

process” OR "business process management”)

We decided not to include words like

"conceptual" and "methodology" since generally, the

approach is not explicitly stated in those terms,

although it clearly appears in a reading of the study.

2.3 Study Selection

To select the studies, inclusion and exclusion criteria

had to be defined, according to the research question

and the search strings defined. The inclusion criteria

were first applied to select a reduced set of relevant

studies to which to apply the exclusion criteria to

select the primary studies, which responded to the

research question. To be selected, the proposals had

to deal with the methodological and/or conceptual

application of SOC and MDD paradigms to business

processes, presenting a methodology, unified

proposal relating the paradigms, transformations and

relations between business processes and service

ICSOFT 2010 - 5th International Conference on Software and Data Technologies

models and notations. Once the articles were filtered

by reading the title, abstract and key words of all of

the articles obtained, the relevant ones were selected

according to the inclusion criteria. The exclusion

criteria were then applied, by reading the articles

thoroughly to select the primary studies with

relevant information about the research issue.

Studies were excluded if they focused on technology

issues, or composition and/or generation of Web

Services (WS) and languages for execution as WS-

BPEL (WS-BPEL) or implementation with

workflows (WfMc), integrated other paradigms such

as agents or grid computing, dealt solely with one

paradigm or did not include business processes as

their main focus.

Table 1: Results obtained for the selected sources.

Source

Found

with

rep.

Relevant

not rep.

Primary

ACM DL 259 3 3

IEEE DL 779 10 8

SCOPUS 264 8 5

Science Direct 503 5 2

Wiley InterScience 8 0 0

Subtotal (2000-2007) 1813 26 18

ACM DL 426 3 0

IEEE DL 300 8 4

SCOPUS 327 11 7

Science Direct 499 3 1

Wiley InterScience 4 0 0

Subtotal (2008-2009) 1556 25 12

Total (2000-2009) 3369 51 30

After eliminating the repeated articles and

reading the title, abstract and keywords of the

resulting articles, 26 were selected as relevant in the

first search, which corresponds to the period 2000 to

2007, and 18 of those were selected as primary

studies. In the review update which corresponds to

the period 2008 to June 2009, 25 studies were

selected as relevant, of which 12 were primary

studies, creating a total of 51 relevant non-repeated

studies from which 30 were selected as primary

studies. It is clear that although the total quantity of

articles found is similar for both periods, the first

result covers a period of eight years, while the

second covers a period of only one and a half year,

demonstrating the increasing interest and work being

done in the subject in recent years. The high number

of total articles found decreased to 1.313 after the

elimination of repeated articles between searches

and sources. The selected primary studies are shown

in the Appendix, Table 2.

2.4 Information Extraction

Once the primary studies were selected, the relevant

information was extracted. To perform this task, a

form consisting of many sections was defined,

including general data such as title, publication and

authors with affiliations and associated country, year

of the study, general study description, and an

indication of which paradigms it investigated.

Several key aspects of paradigm integration were

found in the process of information extraction, for

which the studies were classified; the approaches are

presented and discussed in section 4.

3 ANALYSIS OF THE RESULTS

This section analyzes and discusses the contents of

the selected primary studies in order to extract,

organize and present the relevant information.

3.1 Paradigm Integration

Figure 1 (a) shows the publications by paradigm

integration. It can be seen that of the total number of

studies, nearly half (47%) correspond to the

application of service-oriented and model-driven

development paradigms to business processes, and

the other half is divided between the application of

service orientation to business processes (30%) and

the application of model-driven paradigms to

business processes (23%). These results are

consistent with last year’s tendency to unify the

application of business processes and service-

oriented paradigms to support organizational needs,

and the increase in research in the last few years to

provide automated support for this integration.

3.2 Trends in Publication

Figure 1 (b) shows the trends in publications for

paradigm integration. As it can be seen, the greater

number corresponds to the years 2006 and 2008,

bearing in mind that the data shown for 2009 only

reflect the studies found before July of that year, and

the low number for 2007 could be related to the

growth in technical publications that we were not

interested in. It can be concluded that after a first

period of incipient attention in business process

support by services and model-driven development,

this has gained increasingly more attention not only

APPLICATION OF SERVICE-ORIENTED COMPUTING AND MODEL-DRIVEN DEVELOPMENT PARADIGMS TO

BUSINESS PROCESSES - A Systematic Review

Figure 1: Publications by paradigm integration (a) and by year (b).

Figure 2: Business Process (a), and Service/Software (b) modelling notations used in the studies.

from academia but also from organizations, as real

project applications of the proposals have been

conducted, as presented in some studies.

3.3 Notations Used

Regarding business process and service modelling,

one of the most important issues is the notations

used to specify them. Many efforts have been made

to define notations to support the different views of

software development and business process

integration. Different notations can be used for each

identified need ranging from business process

modelling to execution, simulation, the assessment

of desired and undesired properties for the models,

specification of services, service interaction and

composition. As can be seen in Figure 2 (a), the

most widely used notation for business process

modelling is BPMN (BPMN) (29%) and the most

used notation for service/software modelling is

UML (UML) (38%) shown in Figure 2 (b), the

former being the main standard for business process

modelling, since its adoption by the OMG in 2006,

and the latter the accepted standard for software

development. Other notations also used for business

process modelling are UML (16%) and EPC (EPC)

(9%) (Figure 2 (a)), and for service modelling for

business process execution, WS-BPEL (11%) and

WSDL (9%) (Figure 2 (b)) when service

implementation is via WS.

In the majority of the studies, at least one

notation is used or recommended for business

process and service modelling, although generic

notations are sometimes used or no specific notation

is used at all, usually when the approach is

methodological.

3.4 Type of Case Study

Of the total number of studies, 60% correspond to

examples prepared to show different aspects of the

proposal, generally based on standard business

processes used in organizations, and 13%

correspond to real projects in organizations in which

the proposal has been used in joint projects with

industry. On the other hand, 27% of the studies do

not present a case study, generally when the case

study is mentioned as future work. It can be

ICSOFT 2010 - 5th International Conference on Software and Data Technologies

concluded that although some work has been done

on the real application of the proposals, more is

needed to effectively show the benefits for

organizations.

4 MAIN PRINCIPLES IN

PARADIGM INTEGRATION

In this section the main principles regarding SOC,

MDD and BPM paradigm integration as found in the

systematic review are presented and discussed along

with an illustration from the studies. The presence of

each principle in the total of the selected studies is

shown in Figure 3.

Figure 3: Main principles in paradigm integration.

4.1 Business Process Modelling

One of the most important issues for the support of

business processes (BP) by service orientation (SO)

in a model-driven way is how the organizations

handle their business processes, especially regarding

the explicit modelling in some notations, including

the flow of the activities to be performed, the data

exchanged and the roles involved. There are a

variety of notations that can be used to specify

business processes; each has advantages and

disadvantages depending on the needs of the

organization, the type of BP and the use intended for

the model, as can be seen in Figure 2(a).

Studies modelling BP using BPMN are many,

and include (Liew et. al, 2004) where the BPMN BP

model is annotated with extra information to be used

in transformations. (Tao Tao et. al, 2006) model BP

after the analysis of needed services from system

functionalities. (Henkel et al, 2005) model BP that

are then realized by technical processes using

existing services that match business functionality.

(Gacitua-Decar et al., 2008) uses an enhanced

BPMN by domain model elements and an own UML

profile. (Rychly et al., 2008) models BP with BPMN

then transformed into services diagram. (Thomas et

al., 2008) define an initial conceptual model in EPC

for process design transformed into a conceptual-

technical application model in BPMN for process

configuration. (Oquendo, 2008) use BPMN to model

BP which are then transformed into a PI-ADL

formal language. (Touzi et al., 2009) use BPMN to

model collaborative BP adding a special pool called

Collaborative Information System (CIS), which

mediates between partnered information systems.

UML is also used as in (Mili et al, 2006) to

represent and classify generic BP to be used in

several domains which are then instantiated using a

catalogue of software components. (Zdun et al,

2007) use AD to model BP as a long-running

interruptible process called macroflow and IT-

oriented processes as a short-running transactional

process called microflow, where both defined layers

are above a business application service layer.

(Quartel et. al, 2005) uses AD to model BP mapping

them into an ISDL conceptual model related with

application design and implementation models via

the ISDL models. (Herold et al., 2008) uses AD to

model BP, and a use case diagram and business

structure diagram to model other aspects of the

business view. (Bruckmann et al., 2008) uses AD to

model BP representing functions with actions and

process with activities.

EPC is found in studies such as (Roser et al.,

2006) to specify the BP from which to obtain

services in UML, (Murzek et. al, 2006) and

(Mendling et. al, 2006) who present horizontal

transformations between different notations, the

former between EPC and ADONIS and the latter

between yEPC and YAWL. Other less frequently

used notations are Concurrent Object Oriented Petri

Nets (COOPN) in (Chen et. al, 2006) to model BP in

the design phase at a PIM level and FSM (Finite

state machine) in (Tao Tao et al, 2007) to model

core BP, specifying generic activities and separating

others applicable to a particular usage context. Some

generic notations are used in studies to show

activities, data and the flow of the process, but with

no specific notations.

4.2 Service Oriented Modelling

Another important issue regarding service support

for business processes is the service-oriented

modelling approach. Once the BP are known and

modelled, each defined task from the process and

even the process itself have to be realized by one or

a group of services. Although the flow of the

execution of services can be automatically obtained

APPLICATION OF SERVICE-ORIENTED COMPUTING AND MODEL-DRIVEN DEVELOPMENT PARADIGMS TO

BUSINESS PROCESSES - A Systematic Review

from the BP model, the definition, design, model

and implementation of services is of great

importance for tracking their correspondence to BP

and to existing or new systems providing them.

Services and software modelling is mainly done

using UML, which is shown in Figure 2(b), as in (de

Castro et al., 2006), where the first step is to define

the needed services and from them obtain use cases,

service processes to support them and then generate

the required service composition. (Roser et al.,

2006) model services showing the definition of

services obtained for each proposed architectural

approach (centralized and decentralized broker and

brokerless) from the CIM description of business,

(Zdun et. al, 2007) use AD at a microflow level,

showing the services involved in the IT technical

processes, (Gacitua-Decar et. al, 2008) categorize

services into: business services abstracting activities

on business entities, and technical services

abstracting functionality and data provided by

applications and to manage issues such as security

and messaging. Although the use of the UPMS

profile for services is mentioned it is not shown in

the study. (Rychly et al., 2008) define a service

diagram using an UML profile, specifying interfaces

providing only one functionality, port, service

consumer and provider, along with sequence and

composition service diagrams. (Herold et al., 2008)

use stereotypes such as ServiceAction, which

represents the services provided by the application

layer with core functionality. (de Castro et al., 2008)

models the IS view at PIM level, with business

services to be offered by the system as uses cases,

and functionalities and process needed, (Bruckmann

et al.,2008) defines software modelling with class

diagrams and state diagrams. (Touzi et al.,2009)

defines three views for SOA model: services for

business functionalities, information for data and

messages exchanged between services, and process

for services interaction.

Other used notation are WS-BPEL and WSDL,

as in (Thomas et al., 2008) where WS-BPEL is

derived for execution from BPMN models, (Hu et

al., 2003) uses WSDL to describe services,

(Oquendo, 2008) uses a formal architectural

language PI-ADL for SOA, including orchestration

and choreography and generate WS-BPEL for

process execution. Other less used notations are as in

(Quartel et. al, 2005) where services are modelled

using an ISDL dialect defining components that

provides application services obtained from the

business model, (Cauvet et al., 2008) where business

services are modeled with three parts: profile (goal),

structure (process) and process part (BP) and service

composition resulting in a BP or BP fragment using

a service composite graph, or generic ones such as in

(Tao Tao et. al, 2006) where services are identified

from essential functionalities of the system.

4.3 Model Transformations

Transformations between models used for the

specification of BP and services are one key aspect

of paradigm integration. Many approaches have

been proposed to transform and generate software

models from BP models, where existing languages

can be used to define mappings and transformations,

although new ones or different approaches are also

defined. The OMG Query/Views/Transformations

(QVT) standard and ATL (Jouault et al., 2006) are

the most relevant examples. Transformations not

only make it possible to automatically obtain

elements of a target model from an origin model, but

also to explicitly specify the correspondences

between elements and the semantics involved.

Vertical transformations from one level of

abstraction to another generally applied in a top-

down way can be found in (de Castro et. al, 2006),

where four PIMs are defined to model system

behaviour: user services, extended use case, service

process and service composition, defining mapping

rules that can be completely or partially automated.

In (Chen et al., 2006), embedded process controllers

are generated from BP to be integrated into existing

IS such as ERP via service interfaces, generating

java components. (Quartel et. al, 2005) define

transformations to travel from one ISDL conceptual

model to another, from business to service

implementation. (Mili et. al, 2006) uses a question

approach defining variation points and BP variants

in generic BP which are then mapped to a software

components library of generic business components,

to automatically assemble software systems. In

(Roser et al., 2006), three different architectural

approaches for software systems (centralized and

decentralized broker and brokerless) are derived

from a CIM description of business, establishing

how services for each approach correspond to BP.

(Zdun et. al, 2007) apply transformations

successively based on defined patterns, starting with

the macroflow-microflow pattern which establishes

the conceptual basis and the process-based

integration architecture pattern that guides the

design of an architecture based on sub-layers for the

service composition layer. (Henkel et al., 2005)

propose going from BP models to technical

processes matching existing services by applying

transformation patterns classified with respect to the

ICSOFT 2010 - 5th International Conference on Software and Data Technologies

quality of the transformation. In (Gacitua-Decar et

al., 2008) conceptual transformations are defined

based on the successively application of patterns

from the top to the bottom layer, using graphs for

pattern matching. (Rychly et al., 2008) define two

steps for BP to services transformation: identifying

tasks in BP representing service invocations, then

using a proposed technique that integrates BP and

object modelling into a Business Service Model

(BSM), mediating between business requirements

and implementation. (Herold et al., 2008) go from a

business model (CIM) to an analysis model (PIM),

identifying serviceAction in tasks, then to a Design

model (Architecture specific model, ASM), mapping

services to the target architecture where each

component provides a set of services. (de Castro et

al., 2008) define transformations from a value model

to use case models, defining mapping rules from the

CIM to PIM level between model elements, which

are automated using ATL and tools. (Bruckmann et

al., 2008) use stereotypes actions that map to user

and system functions in a defined metamodel

described in an XML schema as interchange format

and input for transformation engines. (Oquendo et

al., 2008) define mapping between BPM constructs

and PI-ADL for SOA expressions, showing

mappings for a subset of process patterns and

BPMN core elements. (Touzi et al., 2009) define

two types of transformation rules: basic generation

to create elements of the target model, and binding

rules to generate links between them, using ATL.

Horizontal transformations on the same level of

abstraction can be found in (Murzek et al., 2006)

based on control flow patterns (van der Aalst, 2003),

identifies patterns in the original BP, transforming

each one into the target notation to obtain the target

BP model, and in (Mendling et al. 2006) where

transformations are based on elements of each

notation and the algorithm traverses the yEPC

process graph node by node to transform the BP. In

(Sadiq et al., 2006) an automatic distribution of

collaborative BP from an integrated one is proposed,

defining the correspondences and algorithms to

extract the distribute models from the integrated one.

A combination of vertical and horizontal

transformations can be found in (Liew et al, 2004),

where the BPMN BP model is annotated with

information processed by the defined algorithms,

and then transformed into several UML software

artifacts: horizontal from BP to AD, vertical from

BP to use cases, collaboration and deployment

diagrams. (Orriens et al., 2006) define mappings

between defined models in three levels that can be

horizontal or vertical; using five elements capturing

particular facets: what, how, where, who and when.

(Thomas et al., 2008) define horizontal

transformations between EPC conceptual model to

BPMN conceptual-technical model, and vertical

ones from BPMN to BPEL for process execution.

4.4 Methodological Approach

When modelling business processes, services and

other software artifacts needed to support software

development, a systematic approach to guide the

development is essential. Even if some artifacts can

be obtained automatically from others, a guide for

the activities to be done and the flow between them,

among other aspects, are a key factor for success.

Software development processes have been

successfully used in recent years, such as Unified

Process (Jacobson et al., 1999), and approaches to

include service views, activities and artifacts to

guide service development have also been defined.

(Papazoglou et. al, 2006) define a methodology

for SO design and development from business

models, defining SO design and development

principles such as service coupling, cohesion and

granularity. It defines six phases: planning, analysis

(process identification, process scoping, business

gap analysis and process realisation) and design

(service design concerns, specification, business

processes), construction and testing, provisioning

(service governance, certification, metering and

rating, billing strategies), deployment, execution and

monitoring which are traversed iteratively.

(Kohlborn et al., 2009), after reviewing thirty

existing service development approaches, propose a

consolidated approach that combines examined

methodologies and adds new items. They define two

main parts for the process: the derivation of business

services with four phases of preparation,

identification, detailing and prioritization, and the

derivation of software services to support them with

phases: preparation, identification and detailing.

(Tao Tao et. al, 2006) based on the methodology

(Papazoglou et al. 2006) adopts three primary phases

of BP analysis, design and implementation, defining

for BP analysis the steps: services identification,

component identification, process scoping and

process realization analysis. (Zhao et al., 2006)

define the following phases: contracting,

collaboration and design to define services provided

and required by each organisation involved and to

design and coordinate collaborations. (Herold et al.,

2008) propose a model-driven approach with four

phases: business development, requirement analysis,

architectural design and implementation modelling,

APPLICATION OF SERVICE-ORIENTED COMPUTING AND MODEL-DRIVEN DEVELOPMENT PARADIGMS TO

BUSINESS PROCESSES - A Systematic Review

with guidelines and transformations to move from

one model to the other. (de Castro et. al, 2006)

defines a method for service composition with a

process comprising several steps related with model

generation, defining metamodels, models and

artifacts to be obtained from each step, specifying

activities with tasks and inputs and outputs. The

business model is the general input for the process

and its output the services composition model.

(Gacitua-Decar et al., 2008) define steps to apply

pattern techniques to successively refine BP into

services that realize them, identifying business

patterns in BP, and technical services. (Thomas et

al., 2008) define three phases: process design,

configuration and execution including in each the

defined models and transformations. (de Castro et

al., 2008) define an SOD-M service-oriented

development method with an MDA-based approach

with a process and transformations between models

as in (de Castro et al., 2006). (Touzi et al., 2009)

define models, metamodels and transformations to

go from collaborative BP (CIM) to the SOA model

(PIM) from which to generate code (BPEL), based

on the PIM4SOA paradigm.

For B2B development, in (Baghdadi, 2004) a

design process is proposed, consisting of six steps:

BP specification, decomposition and distribution

specification, mapping and validation, supporting

services and components specification, logical B2B

application architecture, implementation and

integration technology, defining objective, input and

output artifacts and models and tools.(Hu et al.,

2003) define steps for going from BP definitions,

implementing activities that can be internal

applications or remote services provided by others,

with a service mediating layer to bridge activity

specifications with its implementation. (Huemer et

al., 2008) define a top-down methodology based on

existing approaches, starting from business and BP

models to services deployment artifacts, outlining a

description, notations and tools to use for each step,

backed with a software factory to generate code.

Other approaches include different visions for

service development, as in (Cauvet et al., 2008),

which defines an iterative service composition

process in which services matching BP requirements

are selected and alternative services can be

generated, using ontologies to match BP

requirements to service goals. (Chen et al., 2008)

define the BITAM-SOA framework for business-IT

alignment extending ATAM, via architecture,

governance and communication, defining three

layers comprising specific modules, using them to

guide a process model for service design,

development and management, which can be top-

down or bottom-up. In (Tao Tao et al., 2007), four

steps in service development are defined based on

core BP determination, adding a usage context

creation for Configurable Context BP (CCBP)

generation and service interface derivation.

4.5 Use of Patterns

Design patterns for software development are well

known and have been used by the community for a

long time, with the reference point being that of GoF

(Gamma et. al, 1995). For business process

modelling, the most relevant work is (van der Aalst

et al., 2003), in which several BP constructions are

defined and analyzed. The importance of reusing the

best existing solutions for known problems is well

established in the Software Engineering literature, so

another key aspect for paradigm integration is the

use of patterns at various stages of development.

Different pattern approaches are presented, as in

(Zdun et al., 2007), where a pattern language with

patterns and primitive patterns is defined for the

integration of BP and technical processes based on

services, which are applied successively in a top-

down way from a macroflow defined process. In

(Henkel et al., 2005), transformation patterns are

defined to be used when going from BP to technical

processes, applying levels of realization and

realization types of BP by using existing services to

match BP in lossfull, constrained, lossless and

exceeded realizable transformations. (Gacitua-Decar

et al., 2008) define business patterns in two types:

process and domain patterns and SOA patterns; a

pattern catalogue organizes them into templates.

(Murzek et al., 2006) use workflow patterns for

the control flow aspect of BP models, as a basis for

the horizontal transformations between different BP

notations. (Oquendo, 2008) also uses process

patterns to map BPMN constructs to PI-ADL

expressions which are iterative and applied to the

original BP for transformations to services in PI-

ADL. In (Roser et al., 2006), the Broker

architectural pattern is used and patterns for service

interaction are modelled in UML collaborations.

Other architectural patterns are also mentioned in

various studies, being the most used the Layers

pattern to define and organize architectural levels.

4.6 Collaborative Processes

The modelling of collaborative processes adds

complexity and coordination requirements to

business process models. However, collaborative

ICSOFT 2010 - 5th International Conference on Software and Data Technologies

process modelling is one of the most needed

activities in organizations, in order to best define the

way to perform their collaborative business with

their partners in a coordinate and beneficial way.

(Orriens et al., 2006) present a Business

Collaboration Context Framework (BCCF),

capturing models in a business collaboration

information model (BCIM) with three levels:

strategic, operational and service level, with

mappings which are then developed and managed,

driven by rules which make it possible to validate

and verify the alignment between model elements.

In (Roser et al., 2006), SO systems realizing

collaborative BP are derived from a business level in

a model and architecture-driven development

perspective (architecture approaches centralized and

decentralized broker and brokerless) as part of the

ATHENA project, where the PIM4SOA comprising

a set of metamodels and tools allows the description

of services and their collaborations at PIM level.

(Zhao et al., 2006) provide support for BP

collaborations of dynamic virtual organizations on

the basis of a service-oriented relative workflow,

adding definitions for IT and BP, where private

information is hidden by wrapping local workflows

into perceivable workflows according to visibility

constraints for defined perceptions. (Touzi et al.,

2009) propose a methodology for developing

collaborative architectures following the MDA

approach from collaborative BP to the SOA model,

adding specific elements for collaborative modelling

as an intermediate collaborative pool in BPMN

models, and collaborative services from it.

For B2B application development, (Baghdadi,

2004) defines a four-layered architecture with four

interrelated abstraction levels: business models and

process, BP decomposition and distribution,

supporting services and integration technology to

guide the design process. (Hu et al., 2003) define a

three-level conceptual framework and architecture to

flexible service enactment in B2B collaborative

processes, with a service-mediating layer to bridge

the BP definition with its implementation with

services. (Huemer et al., 2008) define three layers

based on the Open-edi reference model for inter-

organizational systems: business operational view

(BOV) comprising business models and BP model

layers, and the functional service view (FSV)

comprising a deployment artifacts layer.

4.7 Tool Support

Another key aspect for paradigm integration is tool

support for each development stage in the business

and software development effort. To effectively help

closing the business and systems gap, the use of

tools that enable a smooth integration between both

areas, models and artifacts is needed.

Studies that provide their own tool support

include (Orriens et al., 2006) with the tool ICARUS

developed to support the framework proposal,

(Sadiq et al., 2006) with a BP editor that implements

the algorithms for BP distribution and (Zdun et al.,

2007) with a model-driven tool chain that supports

modelling and model validation, using UML

transformed into DSL syntax, validating models and

transforming them into EMF for generating Java and

BPEL code. (Oquendo, 2008) uses an own core

toolset and customizable tools for PI-ADL

developed previously for the Archware project.

Other studies use existing tools, adding their

own support when needed, as in (Chen et al., 2006)

where an IDE COOPN Builder is used to edit,

visualize and verify COOPN modules, including a

java and WS generator. (de Castro et al., 2006) use

the Oracle BPEL development tool and own tools

developed for the MIDAS framework, while in

(Quartel et al., 2005) an ISDL editor and simulator is

used adding a prototype of a BPEL profile for ISDL.

(Mili et al., 2006) use EMF to define metamodels

and the Eclipse BPEL plug-in to model BP. (Tao

Tao et al., 2006) use the Oracle BPEL process

manager for implementation and EJBs, and

apacheAxis and MySQL as infrastructure. (Roser et

al., 2006) use ARIS for BP modelling and a

prototype implemented in ATHENA for the

generation of services and BPEL processes, (Hu et

al., 2003) uses the IBM MQSeries workflow as a

basis for a prototype implementation of the proposal.

(Huemer et al, 2008) evaluate the MS DSL Tools for

Visual Studio and ADONIS as candidate tools for

the software factory. (de Castro et al., 2008) use

Eclipse with EMF for metamodel definition, GMF

for model visualization and ATL for

transformations, working on code generation into

different WS platforms. In the same direction,

(Touzi et al., 2009) use a set of Eclipse tools: Intalio

designer for collaborative BP modeling, EMF for

metamodel definition, ATL for transformations and

TOPCASEDO for visualizing UML models.

References for the tools can be found in the studies.

5 CONCLUSIONS

This article has presented a systematic review of the

literature concerning the application of service-

oriented and model-driven paradigms to business

processes. The results obtained constitute the

APPLICATION OF SERVICE-ORIENTED COMPUTING AND MODEL-DRIVEN DEVELOPMENT PARADIGMS TO

BUSINESS PROCESSES - A Systematic Review

starting point for research into the methodological

and conceptual approach of paradigm integration,

which beyond a specific implementation or

technology enables the systematic application of the

concepts defined in the paradigms.

The selected studies present different lines of

work and several important aspects of paradigm

integration, which allowed us to extract the main

principles that have to be taken into account when

carrying out paradigm integration. Business process

and service modelling have proven to be two of the

most important principles regarding the relation

between business and software areas. To move from

one area to another, two other principles were found

to provide the basis for bringing the areas closer: the

methodological and the model-driven approach. The

former provides guidelines to develop services

systematically from business processes, while the

latter, by means of mappings, rules and

transformations, makes the automation of

development possible, with models and metamodels

being the key elements. Two important principles to

improve the quality of the solutions are: the reuse of

patterns for BP and service modelling which reduce

errors in early stages of development, and the tool

support for development, including facilities for the

verification of model properties, simulation of

process and execution of transformations.

Finally, it must be noticed that the search was

reduced to a limited number of search engines and

excluded studies which address paradigm integration

but mainly in a purely technological way that did not

provide any contributions in the context of this

work. However, since this is such a broad area and

one of growing interest, future work will include

updates and reviews of other sources to include new

studies that contribute to the ongoing research work.

ACKNOWLEDGEMENTS

This work was partially funded by the Agencia

Nacional de Investigación e Innovación (ANII) from

Uruguay, ALTAMIRA project (Junta de

Comunidades de Castilla-La Mancha, Spain, Fondo

Social Europeo, PII2I09-0106-2463) and

PEGASO/MAGO project (Ministerio de Ciencia e

Innovación MICINN, Spain, and Fondo Europeo de

Desarrollo Regional FEDER, TIN2009-13718-C02-

01).

REFERENCES

Baghdadi Y., 2004.ABBA: an architecture for deploying

business-to-business electronic commerce

applications, In Electronic Commerce Research and

Applications, Vol. 3, Is. 2.

(BPMI), Business Process Management Initiative <http://

www. bpmi.org/>

Bruckmann T., Grunh V., 2008. AMABULO- A Model

Architecture for Business Logic, In ECBS’08, 15th

IEEE International Conference and Workshop on the

Engineering of Computer Based Systems.

(BPMN) Business Process Modeling Notation, Object

Management Group (OMG), 2006 - 2008.

Cauvet C., Guzelian G., 2008. Business Process Modeling:

a Service-Oriented Approach, In HICSS’08, 41st

Hawaii International Conference on System Sciences.

Chen, A., Buchs, D., A., 2006. Generative Business

Process Prototyping Framework, In RSP’06, 16th

International Workshop on Rapid System Prototyping.

Chen H., 2008. Towards Service Engineering: Service

Orientation and Business-IT Alignment, In HICSS’08,

41st Annual Hawaii International Conference on

System Sciences.

De Castro, V., Marcos, E., López Sanz, M., 2006. A

model driven method for service composition

modelling: a case study. In International Journal Web

Engineering and Technology, Vol. 2, No. 4.

De Castro V., Vara Mesa J. M., Herrmann E., Marcos E.,

2008. A Model Driven Approach for the Alignment of

Business and Information Systems Models, In ENC

'08, 9th Mexican International Conference on

Computer Science.

(EPC) van der Aalst, W, 1999, Formalization and

Verification of Event-driven Process Chains.

Information and Software Technology 41, 639–650.

Erl, T., 2005. Service-Oriented Architecture: Concepts,

Technology, and Design, Prentice Hall.

Gamma, E., Helm, R., Johnson, R., Vlissides, J., 1995,

Design Patterns: Elements of Reusable Object-

Oriented Software, Addison-Wesley

Gacitua-Decar V., Pahl C., 2008. Pattern-based business-

driven analysis and design of service architectures, In

ICSOFT‘08, 3rd International Conference on

Software and Data Technologies.

Henkel, M., Zdravkovic, J., 2005. Supporting

Development and Evolution of Service-based

Processes, In ICEBE’05, 2nd. International

Conference on e-Business Engineering.

Herold S., Rausch A., Bosl A., Ebell J., Linsmeier C.,

Peters D., 2008. A Seamless Modeling Approach for

Service-Oriented Information Systems, In ITNG’08,

5th International Conference on Information

Technology:New Generations.

Hu J., Grefen P., 2003. Conceptual framework and

architecture for service mediating workflow

management, In Information and Software

Technology, Vol.45, Is. 13.

ICSOFT 2010 - 5th International Conference on Software and Data Technologies

Huemer C., Lieg P., Schuster R., Werthner H., Zapletal

M., 2008. Inter-organizational systems: from business

values over business processes to deployment, In

IEEE-DEST’08, 2nd IEEE International Conference

on Digital Ecosystems and Technologies.

Jacobson, I., Booch, G., Rumbaugh, J., 1999, The Unified

Software Development Process, Addison Wesley.

Jouault, F., Kurtev, I., 2006, Transforming Models with

ATL (ATLAS Transformation Language), Satellite

Events at the MoDELS 2005 Conference.

Kitchenham, B., 2004. Procedures for Performing

Systematic Reviews. Software Engineering Group,

Department of Computer Science, Keele University

and Empirical SE National ICT, Australia.

Krafzig, D. Banke, K. Slama, D., 2005. Enterprise SOA,

Service Oriented Architecture Best Practices, Prentice

Hall.

Kohlborn T., Korthaus A., Chan T., Rosemann M., 2009.

Identification and Analysis of Business and Software

Services- A Consolidated Approach, In IEEE

Transactions on Services Computing, Vol. 2, No. 1.

Liew, P., Kontogiannis, K. Tong, T., 2004. A Framework

for Business Model Driven Development. In STEP’04,

12th International Workshop on Software Technology

and Engineering Practice.

(MDA), Model Driven Architecture, 2003, Guide version

1.0.1, Object Management Group (OMG)

Mellor, S., Clark, A., Futagami, T., 2003. Model Driven

Development - guest editors' introduction, In IEEE

Computer Society, September/October edition.

Mendling, J., Moser, M., Neumann, G., 2006.

Transformation of yEPC Business Process Models to

YAWL. In SAC’06, 21st. Symposium on Applied

computing.

Mili, H., Frendi, M., et. al, 2006. Classifying Business

Processes for Domain Engineering, In ICTAI’06, 18th

International Conference on Tools with Artificial

Intelligence.

Murzek, M., Kramler, G., Michlmayr, E., 2006. Structural

Patterns for the Transformation of Business Process

Models, In EDOCW’06, 10th International Enterprise

Distributed Object Computing Conference Workshops.

Oquendo F., 2008. Formal Approach for the Development

of Business Processes in terms of Service-Oriented

Architectures using PI-ADL, In SOSE’08, 4th. IEEE

International Symposium on Service-Oriented System

Engineering.

Orriens, B., Yang, J., Papazoglou, M., 2006. A Rule

Driven Approach for Developing Adaptive Service

Oriented Business Collaboration, In SCC’06, 3rd.

International Conference on Services Computing.

Papazoglou, M., van den Heuvel, W., 2006. Service-

oriented design and development methodology.

International Journal Web Engineering and

Technology, Vol. 2, No. 4.

Papazoglou, M., Traverso, P., Dustdar, S., Leymann, F.,

2007. SOC: State of the Art and Research Challenge,

In IEEE Computer Society, November edition.

Quartel, D., Dijkman, R., van Sinderen, M., 2005. An

approach to relate business and application services

using ISDL, In EDOC’05, 9th International

Enterprise Computing Conference.

(QVT) Query/Views/Transformations, 2008, v. 1.0, Object

Management Group (OMG)

Rychly M., Weiss P., 2008. Modeling of Service Oriented

Architecture: from business process to service

realization, In ENASE’08, 3rd International

Conference Evaluation of Novel Approaches to SE.

Roser, S., Bauer, B., Muller, J., 2006. Model- and

Architecture-Driven Development in the Context of

Cross-Enterprise Business Process Engineering, In

SCC’06, 3rd. Int. Conference on Services Computing.

Sadiq, W., Sadiq, S., Schulz, K., 2006. Model Driven

Distribution of Collaborative Business Processes, In

SCC’06, 3rd. Int. Conference on Services Computing.

Smith, H., Fingar, P., 2003. Business Process

Management: The third wave, Meghan-Kieffer Press

Stahl, T., Volter, M. et. al., 2006, Model-Driven Software

Development, Technology, Engineering, Management,

John Wiley & Sons, Ltd.

Tao Tao, A., Yang, J., 2006. Develop Service Oriented

Finance Business Processes: A Case Study in Capital

Market, In SCC’06, 3rd. International Conference on

Services Computing.

Tao Tao, A., Yang, J., 2007. Supporting Differentiated

Services With Configurable Business Processes, In

ICWS’07, 5th. International Conference on WS.

Thomas O., Leyking K., 2008. Using Process Models for

the Design of Service-Oriented Architectures:

Methodology and E-Commerce Case Study, In

HICSS’08, 41st Annual Hawaii International

Conference on System Sciences.

Touzi J., Benaben F., Pingaud H., Lorré J. P., 2009. A

model-driven approach for collaborative service-

oriented architecture design, International Journal of

Production Economics, Vol. 121, Is. 1.

(UML) Unified Modeling Language, Object Management

Group (OMG), 2005.

Van der Aalst, W., ter Hofstede, A., Kiepuszewski, B.,

Barros, A. 2003. Workflow Patterns, In Distributed

and Parallel Databases, 14(3), pages 5-51.

(WS), Web Service Architecture, World Wide Web

Consortium(W3C)<http://www.w3.org/TR/ws-arch/>

(WfMC), Workflow Management Coalition, <http://

www.wfmc.org/>

(WS-BPEL) WS- Business Process Execution Language,

Advancing Open Standards for the Information

Society (OASIS) <http://www.oasis-pen.

org/ committees/tc_home.php? wg_abbrev= wsbpel>

Zdun, U., Hentrich, C., Dustdar, S., 2007. Modeling

Process-Driven and Service-Oriented Architectures

Using Patterns and Pattern Primitives, ACM

Transactions on the Web, Vol. 1, No. 3, Article 14.

Zhao, X., Liu, C., Yang, Y., 2006. Supporting Virtual

Organisation Alliances with RelativeWorkflows, In

APCCM’06, 3rd. Asia-Pacific Conference on

Conceptual Modelling.

APPLICATION OF SERVICE-ORIENTED COMPUTING AND MODEL-DRIVEN DEVELOPMENT PARADIGMS TO

BUSINESS PROCESSES - A Systematic Review

APPENDIX

The selected primary studies in the systematic review are presented in Table 2.

Table 2: Primary studies selected.

Id Title Authors/Year Paradigms integration

1 A Framework for Business Model Driven Development Liew et al, 2004 BP+MDD

2 Generative Business Process Prototyping Framework, Chen et al, 2006 BP+SOC+MDD

A model driven method for service composition modelling: a case

study

de Castro et al, 2006 BP+SOC+MDD

A Rule Driven Approach for Developing Adaptive Service Oriented

Business Collaboration

Orriens et al, 2006 BP+SOC+MDD

5 An approach to relate business and application services using ISDL Quartel et al, 2005 BP+SOC+MDD

6 Classifying Business Processes for Domain Engineering Mili et al, 2006 BP+MDD

Develop Service Oriented Finance Business Processes: A Case Study

in Capital Market

Tao Tao et al, 2006 BP+SOC

Model- and Architecture-Driven Development in the Context of

Cross-Enterprise Business Process Engineering

Roser et al, 2006 BP+SOC+MDD

9 Model Driven Distribution of Collaborative Business Processes Sadiq et al, 2006 BP+MDD

Modeling Process-Driven and Service-Oriented Architectures Using

Patterns and Pattern Primitives

Zdun et al, 2007 BP+SOC+MDD

11 Service-oriented design and development methodology Papazoglou et al, 2006 BP+SOC

Structural Patterns for the Transformation of Business Process

Models

Murzek et al, 2006 BP+MDD

13 Supporting Development and Evolution of Service-based Processes Henkel et al, 2005 BP+MDD

Supporting Differentiated Services With Configurable Business

Processes

Tao Tao et al, 2007 BP+SOC+MDD

15 Supporting Virtual Organisation Alliances with RelativeWorkflows Zhao et al, 2006 BP+SOC

16 Transformation of yEPC Business Process Models to YAWL Mendling et al, 2006 BP+MDD

ABBA: an architecture for deploying business-to-business electronic

commerce applications

Baghdadi, Y. 2004 BP+SOC

Conceptual framework and architecture for service mediating

workflow management

Hu et. al, 2003 BP+SOC

Pattern-based business-driven analysis and design of service

architectures

Gacitua-Decar el al, 2008 BP+SOC+MDD

Inter-organizational systems: from business values over business

processes to deployment

Huemer et. al, 2008 BP+SOC

Modeling of Service Oriented Architecture: from business process to

service realization

Rychly et. al, 2008 BP+SOC+MDD

22 Business Process Modeling: a Service-Oriented Approach Cauvet et. al, 2008 BP+SOC

Using Process Models for the Design of Service-Oriented

Architectures: Methodology and E-Commerce CaseStudy

Thomas et. al, 2008 BP+SOC+MDD

A Seamless Modeling Approach for Service-Oriented Information

Systems

Herold et. al, 2008 BP+SOC+MDD

A Model Driven Approach for the Alignment of Business and

Information Systems Models

de Castro et al, 2008 BP+SOC+MDD

Towards Service Engineering: Service Orientation and Business-IT

Alignment

Chen, H., 2008 BP+SOC

Identification and Analysis of Business and Software Services- A

Consolidated Approach,

Kohlborn et. al,2009 BP+SOC

28 AMABULO- A Model Architecture for Business Logic Bruckmann et. al, 2008 BP+MDD

Formal Approach for the Development of Business Process in terms

of Service-Oriented Architectures using PI-ADL

Oquendo, F., 2008 BP+SOC+MDD

A model-driven approach for collaborative service-oriented

architecture design

Touzi et. al, 2009 BP+SOC+MDD

ICSOFT 2010 - 5th International Conference on Software and Data Technologies