Extrinsic Dependencies in Business Process Management Systems

Radhwan Mahdi, Stefan Jablonski and Stefan Sch

onig

Institute for Computer Science, University of Bayreuth, Germany

Keywords:

Process Management, Context Awareness, Extrinsic Dependencies.

Abstract:

The demand for supporting the ﬂexibility in business processes has been increasing due to dynamic business

environments and technological progress. That led to the challenge of designing business processes so as

to take context changes into consideration. A context refers to any circumstance of a process and includes

factors which impact process execution steps. To overcome this challenge and better ﬁt business processes to

customers expectations, this paper conceptualizes contextual factors relevant to the business process descrip-

tion. It deﬁnes a model that explains how the relevant contextual factors could be identiﬁed and computed in

a structured way. To verify the applicability of the identiﬁed approach, a prototype is set up for running the

experiments. It examines the approach with real information in different real-life scenarios.

1 INTRODUCTION

Over the last two decades, business process mana-

gement (BPM) has been considered as an essential

approach to achieve high performance of organizati-

ons and improve customer satisfaction (Dumas et al.,

2013) (Weske, 2012). BPM is a comprehensive ap-

proach designed to achieve business objectives by

managing end-to-end processes in a structured way

(Hammer, 2010). Modelling business processes pro-

perly is an essential step for the BPM lifecycle. The

traditional modelling approach relies on describing

processes by ﬁve fundamental perspectives (Jablonski

and Bussler, 1996): the functional perspective (what

are the process steps?), the data perspective (what

data consumed or produced?), the operational per-

spective (which tools are used?), the organizational

perspective (who is performing the process?), and the

behavioral perspective (when should activities be per-

formed?). This approach meets the expectations in a

given context (Saidani and Nurcan, 2007). However,

during the last decade, companies have increasingly

been confronted with ﬂexibility challenges like adjus-

ting their business processes to changing circumstan-

ces to best meet customer’s expectations. The main

reasons for that are (Saidani and Nurcan, 2007): (i) ra-

pidly changing business environments; (ii) companies

have been increasingly running more complex busi-

ness processes; and (iii) customers expect services to

become more efﬁcient.

Accordingly, process ﬂexibility has to be enhan-

ced, especially when having further development of

information technology and market globalization in

mind (Tao et al., 2008). Flexibility is the ability to

adapt to ambient changes (Soffer, 2005) (Burkhart

and Loos, 2010) (Regev and Wegmann, 2005), in par-

ticular, context changes. As a consequence, ﬂexible

process modelling requires additional information, in

particular, context information. A context refers to

any circumstance of a process. It includes factors

that have an impact on process design and execution

such as location, weather, and policies (Rosemann

and Recker, 2006). The important question w.r.t. ta-

king the contextual information into consideration is

how to identify and derive contextual factors. This

leads to other questions such as what their main cha-

racteristics are. The contribution of this paper is to

deﬁne a model which conceptualizes a process con-

text. The focus of this conceptualization lies on the

contextual factors that are derived from the funda-

mental process perspectives introduced above. This

objective is achieved by deﬁning a methodology for

the approach of deriving contextual factors. The met-

hodology distinguishes between dependencies of de-

signing and executing processes by discerning intrin-

sic and extrinsic dependencies. The approach can be

used in different scenarios like in the development of

technologies such as web services, mobile devices,

and sensors. Capturing contextual information – such

as detecting the current agent’s geographical location

or agent’s device – is thereby facilitated (Modafferi

et al., 2005) (Zhu et al., 2016). Therefore, the ap-

proach at hand offer the possibility to design proces-

ses that are adaptable to contextual changes as well

632

Mahdi, R., Jablonski, S. and Schönig, S.

Extrinsic Dependencies in Business Process Management Systems.

DOI: 10.5220/0006687006320639

In Proceedings of the 20th International Conference on Enterprise Information Systems (ICEIS 2018), pages 632-639

ISBN: 978-989-758-298-1

as the accurate assignment of process steps to partici-

pants at runtime. A prototypical implementation and

it’s application in two case studies shows the practical

applicability of the work at hand.

The structure of this paper as follows: Section 2 il-

lustrates the motivating examples. Section 3 presents

related work. In Section 4 process dependencies are

described. Section 5 explains the methodology of ex-

trinsic dependencies. Section 6 shows the case studies

and introduces a prototype to evaluate the identiﬁed

approach and Section 7 ﬁnally concludes the paper.

2 RUNNING EXAMPLES

Two simple real-life scenarios will illustrate how the

endogenization of contextual changes can improve

business process design and execution.

2.1 Emergency Car Repairs

Consider the process of having your car repaired. Or-

dinarily, drivers would call their insurance company

to solve their car problems. The driver provides some

information about the car such as the insurance num-

ber, the car type, and the location of the car. Upon

claim receipt, the company opens a task to have the

car repaired at a speciﬁc location. This task needs

to be assigned to an eligible technical team. Finally,

the technical team writes a report for the issue after

solving it. In the execution steps, a process needs to

deﬁne a resource and ﬁnd a team to do the task.

In a normal execution procedure, all of the availa-

ble eligible technical teams will be notiﬁed that they

are able to perform the task. One of the possibilities

of optimizing process execution is to ﬁnd the most

eligible team to do the process. This can be achieved

by classifying the eligibility of the teams. The con-

textualized eligibility can be ascertained by data on

team availability, experience level, and location. By

assuming that all the eligible teams are available and

have the same level of experience when they are loca-

ted in different locations, however, process execution

could be optimized by location-sensitive task assign-

ment. In this case, the nearest team could be notiﬁed

as a highly recommended team while all other teams

would be less recommended. The optimized proce-

dure will make the execution more ﬂexible. Moreo-

ver, it reduces customers’ waiting time.

2.2 Remote Conﬁguration of Device

Network companies are divided into different depart-

ments such as a call center, a network operations

center (NOC), and an operations maintenance cen-

ter (OMC). Generally, the call center sends a ticket

to the other departments after a call from a customer.

One of the issues posed to the NOC is to reconﬁgure

a network device remotely. This requires access to

remote network devices via the internet. Due to se-

curity risks, a network engineer has to be authorized

to access these devices. Simple authorization steps

are used when an engineer works from the company

ofﬁce because he is working from an authorized IP

domain. A more complex procedure of authentica-

tion is applied to an engineer who works outside the

company network domain (e.g. home ofﬁce). This

procedure requires additional checking steps to avoid

intrusion by unauthorized persons such as hackers.

Uncontextualized BPMS assigns the process to any

eligible network engineer’s task list, irrespective of

their location. With context-aware BPMS, an indu-

ced location preference, which can be used for loca-

tion optimization, needs to simultaneously use a loca-

tion sensitive authorization procedure that discrimina-

tes between company ofﬁce and home ofﬁce for dif-

ferent authorization steps. Company-ofﬁce engineers

who go through the simpler authorization protocol are

strictly preferred to outside engineers. Additionally,

the preference could be made based on other factors,

for example, the quality of internet connection of the

engineer’s device which is an important aspect to en-

sure the success of the process. Therefore, taking

context information into account will improve process

assignment step by creating different levels of prefe-

rence between agents.

3 RELATED WORK

This section presents and discusses literature which

is strongly related to the goal and methodology of

this paper from different points of view. Mainly, it

covers the context-awareness and location-awareness

in BPM domain. (Modafferi et al., 2005) suggests a

workﬂow application solution to the provision of cu-

stomized services based on user context, one form of

context awareness. Key to their methodology for the

development of context-sensitive business processes

is the concept of the context-sensitive region. This

concept enables the process parts, the behavior of

which may vary with the context, to react to a con-

text change. The research presented in (Rosemann

and Flender, 2007) focuses on the extrinsic drivers for

process ﬂexibility instead of intrinsic ways of adapta-

tion to environmental changes. Another paper (Sai-

dani and Nurcan, 2007) proposes a taxonomy of the

context related knowledge into four kinds as an initial

Extrinsic Dependencies in Business Process Management Systems

633

stage: ”location-related context”, ”time-related con-

text”, ”resource-related context”, and ”organization-

related context”. They introduce the ”context model”

to structure the contextual information. It consists of

three dimensions: 1) aspects are used to capture the

context, 2) facets are used to address each aspect, 3)

each facet is described by attributes which could be

measured. In the above-mentioned researches, the lo-

cation is considered as one of the important factors

in the process context. In (Rosemann and Flender,

2007), the location is considered as an important va-

riable in the environmental context layer of the on-

ion model. In (Saidani and Nurcan, 2007), the loca-

tion is presented as one of the four aspects of con-

text. However, the importance of geospatial informa-

tion in the execution of processes has increased due

to the technological progress (Zhu et al., 2016). In

(Sch

onig et al., 2014), the location is treated as a new

perspective of the process in addition to its fundamen-

tal perspectives. This paper introduces the Declara-

tive Process Intermediate Language (DPIL) (Sch

onig

et al., 2017) for modelling business processes. In or-

der to take the locational information into account, the

meta-model of DPIL is extended to model locational

entities and constraints.

4 DEPENDENCIES IN BPM

Any business process relies on some information for

its design and execution. This information should

capture the important aspects of the process such as

the process steps and the time and order of their exe-

cution. In addition to that, any optimization of the

design or execution relies on some other information.

For example, contextual information is needed to sup-

port ﬂexibility in processes (Rosemann and Recker,

2006)(Saidani and Nurcan, 2007). Here, a distinction

between the required information for design and exe-

cution of the process and for the optimization is nee-

ded to analyze the applicability of context for incre-

asing ﬂexibility. This distinction should facilitate the

conceptualization of the process context and provide

a structured way of deriving the contextual factors.

The process dependencies are divided into intrinsic

and extrinsic dependencies in the following two sub-

sections.

4.1 Intrinsic Dependencies

To complete a BPM lifecycle, information about a

process description is required. Each phase of the li-

fecycle needs speciﬁc information that must suit the

respective phase. In principle, the ﬁve fundamental

perspectives of a process provide the required infor-

mation for these phases. This information provides

a description of a process. Therefore, each process

should have the basic information from these ﬁve per-

spectives. To illustrate that, take an ”oral exam” pro-

cess as a simple example. First, the student needs

to register for the exam. Upon registration, the stu-

dent is given a conﬁrmation notice and the date of the

exam. The second step is to conduct the oral exam.

This step needs to pre-specify the examiner’s name

and room number (normally, the examiner’s room).

Finally, the examiner evaluates the student’s answers

and gives him or her a ﬁnal grade. If executed nor-

mally, the execution depends on the information from

the ﬁve perspectives as a basis. One of the execution

steps is specifying an organization for the process. It

requires general information about the agent such as

the agent’s group. Normally, there are two groups

who are able to play the examiner role: the group of

the professors and the group of the professors’ assis-

tants. Further information, such as the location of the

agent, is not required in this execution. Therefore,

some information from these perspectives is conside-

red as the mandatory information for designing and

executing the process. Here, this mandatory informa-

tion is referred as the intrinsic dependencies of the

process. According to this understanding, the intrin-

sic dependencies can be deﬁned as Intrinsic depen-

dencies: the minimum information which is required

to describe the fundamental perspectives of a process

for the purpose of enabling the process designing and

executing

4.2 Extrinsic Dependencies

As explained in the previous section, a comprehen-

sive description of a process depends on the infor-

mation provided by the ﬁve fundamental perspecti-

ves. Some of this information has been considered

as mandatory information. That raises the questions

of what other information the fundamental perspecti-

ves provide, how to exploit it, and in which step the

process depends on this information. Generally, this

information provides many details about the process

that could be used for execution optimization. To il-

lustrate that point, two simple scenarios of optimizing

the execution of the ”oral exam” process (which pre-

sented in the previous subsection) are given as fol-

lows: (i) The ﬁrst scenario is about the step of as-

signing the process to an examiner. Generally, the

process would be assigned either to the professor or

the professor’s assistants. In case of an emergency

at the exam date, the exam could be postponed. One

of the possibilities to optimize this process is to re-

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

634

assign the exam to somebody who is able to play the

examiner role at that date. This optimization needs

some other information such as the availability of ot-

her agents and their location; (ii) The second scenario

concerns the room assignment for the oral exam. Nor-

mally, any available room inside the faculty could be

selected as a place for the exam. This step is optimi-

zed by taking other information such as the location

of the room into account. Thereby, the room nearest

to the examiner’s ofﬁce would be chosen in an optimi-

zed execution. Some other information such as room

size and the number of chairs might be important in

some other cases. In these scenarios, it is clear that

some information of the fundamental perspectives is

not directly related to the process. Therefore, this in-

formation is not necessary for designing and execu-

ting the process. Nevertheless, it might have an im-

pact on the execution steps. As a consequence, this

information could be connected to the mandatory in-

formation (intrinsic dependencies) and create objects

that make an impact on the process execution steps.

Here, these impacts are considered as the extrinsic de-

pendencies of a process. Based on this understanding,

the extrinsic dependencies are deﬁned in this paper as

Extrinsic dependencies: the information required to

improve the execution of a process on the condition

that it is produced by the entanglement of objects of

this process according to speciﬁc requirements.

4.3 Context Classiﬁcation

The notion of context has been deﬁned many times

across different disciplines and subdisciplines. In the

BPM domain, some research has been published on

developing context-aware business processes. Lea-

ding up to the deﬁnition of a business process con-

text used in this paper, the other deﬁnitions in this

domain are presented ﬁrst. In (Rosemann and Rec-

ker, 2006), business process context has been deﬁned

as ”the minimum set of variables containing all rele-

vant information that impact the design and execution

of a business process”. (Saidani and Nurcan, 2007),

instead, focuses on the impact of context on the as-

signment relations, so the context is deﬁned as ”the

collection of implicit assumptions that is required to

activate accurate assignments in the BP model at the

process instance level”. In a third approach, (de la

Vara, Jose Luis et al., 2010) speciﬁes it as the envi-

ronment properties in which the business processes

are executed. A context is characterized as 1) a set

of states, 2) a subjective, and 3) a relative. It refers

to the environment of the process and is deﬁned as

”any information that is relevant to and might affect

the execution of a business process”.

Taking the above-mentioned views as a basis, we

can subsume the meaning of a business process con-

text with regard to our area of focus. A context in-

cludes information that can refer to any circumstance

of a business process (for example, ”the weather is

hot”, ”the agent is not available”). This information

constitutes factors called context factors. For exam-

ple, information about the agent’s occupancy status

(busy or available) can form a factor called ”availabi-

lity”. These factors could impact the process execu-

tion steps such as the assignment of a task to an agent.

Generally, these factors can be captured and measu-

red by variables which are called context variables.

The values of these variables vary with the changes

in circumstances of a process (Rosemann and Recker,

2006). As presented in the example ”repairing a car”

in section 2.1, one of the possibilities to improve the

execution of this process is to assign the task to the ge-

ographically nearest team. To do that, the context va-

riable ”distance” can be used to measure the distance

from each team’s location, which is speciﬁed in the

form of latitude and longitude (Imagery and Agency,

2000), to the location of the event. Comparing the va-

lues of the ”distance” variable, the nearest team could

be speciﬁed.

According to the above understanding of business

process context, we could deﬁne it in accordance with

the previous deﬁnitions as a global (general) term

which consists of factors impacting the business pro-

cess execution steps. These factors could be deri-

ved from information about the process description or

process environment changes. The main focus of this

paper is on the contextual factors that are related to the

process internally (i.e., they are related to the process

perspectives), in particular, organizational and opera-

tional perspectives are of most interest here. These

factors have a direct impact on the process execution.

In addition to that, they can be captured and measured

accurately, especially given the development of mo-

bile technologies. According to our above-mentioned

deﬁnition of process context, the context factors can

be classiﬁed into two types. The ﬁrst one is related

to the process itself and the second type is related

to the environment in which the process is executed.

The classiﬁcation is explained as follows: Extrinsic

factors: They include elements that can be compu-

ted from the intrinsic objects of a process (i.e., they

are related to the fundamental perspectives). In other

words, the process description is the main source for

deriving these elements. These factors make a direct

impact on the process execution steps. As explained

in section 4.2, this impact is considered as the extrin-

sic dependencies of a process. Environmental fac-

tors: They refer to any information outside the pro-

Extrinsic Dependencies in Business Process Management Systems

635

cess construction (i.e. they are not related to the pro-

cess description). Generally, it covers all the elements

of the system in which the process is executed. Ele-

ments come from the organization side (such as cu-

stomers, competitors, strategy) and the environment

(such as weather and culture). This type could be

classiﬁed into some other types (based on (Rosemann

and Flender, 2007)) but the current classiﬁcation suf-

ﬁces for this research. Based on the above-mentioned

area of focus, the derivation of the extrinsic factors is

provided in the next section. Additionally, an appro-

priate model for the conceptualization of the factors

is set up.

5 DEPENDENCIES MODELLING

In this section, a detailed approach for the concep-

tualization of a business process context is presented

based on the deﬁnition and classiﬁcation of context

in Sec. 4.3. A conceptual model of the formation (or

derivation) of contextual factors is provided.

5.1 Process Objects

Based on the concept expounded in (Jablonski and

Bussler, 1996), every process has a number of rele-

vant artifacts. These artifacts are modeled as objects.

In addition to that, an object type is used to model a

single artifact type. The object has properties which

are used to provide speciﬁcations of the object. By

following this concept, the process constitutes a set

of the relevant objects refer to its description (i.e. the

ﬁve fundamental perspectives). Generally, the objects

could also belong to any other perspective that des-

cribes the process such as causality, integrity, quality,

history, and security.

In this paper, any object that describes the pro-

cess internally (i.e., it is related to the fundamental

perspectives) is called an intrinsic object. The impor-

tant question is which kind of relation relates the in-

trinsic objects to the process. Depending on our dis-

tinction between process dependencies, there is some

information which is mandatory to design and execute

the process whereas some other information might be

required for further steps such as optimization. As

a consequence, there are some objects mandatory to

complete process description. These objects are di-

rectly related to the process perspectives. For exam-

ple, an agent is an object which directly describes the

organizational perspective of a process. In this paper,

these kinds of objects are called direct objects. By

contrast, there are some other objects which do not di-

rectly describe the perspectives. Normally, these ob-

jects are connected to the direct objects. That means

they are indirectly related to the process. These ob-

jects are called the indirect objects of the process. For

example, the agent’s device is an object connected to

the agent which is a direct object of a process. In

short, the process objects might be directly or indi-

rectly related to the process. These types are deﬁned

as follows: Direct objects: the set of objects manda-

tory for the design and execution of the process which

directly describes the process perspectives. Indirect

objects: an additional set of objects that is connected

to a speciﬁc direct object to provide additional infor-

mation about the process. In summary, the intrinsic

objects of a process include any information that re-

fers to the process perspectives. These objects can

be divided into direct and indirect objects, in accor-

dance with the relation of the information provided

to the process. While the direct objects are necessary

for process lifecycle phases, the information provided

by indirect objects is useful for other steps such as

optimization. The division of process objects facilita-

tes the computation of the context variables values as

presented in the subsequent subsections.

5.2 Formation of Extrinsic Factors

Extrinsic factors greatly inﬂuence different process

execution steps, especially, as they relate to process

perspectives (cf. Sec. 4.3 for the classiﬁcation of pro-

cess context). The impact of these factors is consi-

dered as the extrinsic dependencies of a process (cf.

Sec. 4.2 for deﬁnitions). Therefore, identifying and

computing these factors is one of the important steps

of modelling extrinsic dependencies. Generally, ex-

trinsic factors include elements that can be derived

from the intrinsic objects of a process. In other words,

the direct and indirect objects are intertwined to cre-

ate new information. This information is represented

in variables which are called extrinsic variables (or

context variables).

At the end of this step, questions about the main

characteristics of extrinsic factors and their derivation

have to be answered. For that, a well-deﬁned proce-

dure which is able to derive the extrinsic factors and

compute its variables should be followed. Basically,

two main steps can be executed as a procedure to form

the extrinsic factors: 1) identiﬁcation and 2) compu-

tation of the extrinsic variables. The procedure is ex-

plained as follows: Identifying the extrinsic factors:

This step is responsible for identifying the relevant ex-

trinsic factors that have an impact on the execution of

a process. This identiﬁcation needs information about

the required optimization. Normally, this information

is provided by the context requirements. The deter-

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

636

Fig. 1: Process dependencies model.

mination of context requirements strongly depends on

the objectives of a process. Therefore, it differs from

scenario to scenario. Generally, it is deﬁned as infor-

mation that indicates the required elements for a spe-

ciﬁc optimization. Computing the extrinsic varia-

bles: This step speciﬁes the variables of the relevant

extrinsic factors that have been identiﬁed in the ﬁrst

step. Here, these variables are called the extrinsic va-

riables (which are part of the context variables). The

values of these variables change in accordance with

the changes in the related circumstances. To compute

these values, the intrinsic objects of a process (both

direct and indirect objects) are intertwined with each

other. This computation is strongly based on the re-

quirements of the relevant extrinsic factors.

To illustrate the above procedure, a simple sce-

nario of optimizing the repairing a car process (cf.

Sec. 2.1) is discussed. This process can be optimi-

zed by using various pieces of contextual informati-

ons such as location, availability, and the teams’ capa-

bilities. By taking only the location information into

account, the optimization can be achieved by assig-

ning the task according to the team’s location. The

procedure of forming the extrinsic factors is applied

as follows: (i) The context requirement of this opti-

mization is to ﬁnd the nearest team to the location of

the car. According to this requirement, ”the accessibi-

lity” could be identiﬁed as the extrinsic factor for this

optimization; (ii) The variable for the ”accessibility”

factor is the ”distance” where its value varies with the

change of the team’s location. To compute this value,

the geographical location for the event and the team

should be available. According to these coordinates,

the distance is measured.

5.3 Process Dependencies Model

A process model focuses on capturing the elements

that are related to the control ﬂow. That means it sup-

ports processes with predictable paths (Sch

onig et al.,

2014) (i.e., it does not comprise the context elements).

Thus, it focuses only on the intrinsic dependencies of

a process. In keeping with the methodology of this pa-

per, the contextual factors need to be integrated into a

process model. A model to conceptualize the process

context is introduced (depicted by Fig. 1). This mo-

del clariﬁes the ability to take the context factors into

account. The extrinsic and environmental factors are

considered as subclasses of context. In addition, the

identiﬁed procedure for the formation of the extrin-

sic factors is fully illustrated. Intrinsic dependen-

cies part: It refers to the information mandatory for

the design and execution of the process. Hence, this

part of the model covers the direct objects of the pro-

cess and their properties (i.e. they are directly related

to the fundamental process perspective; see section

5.1 for more detail). Extrinsic dependencies part:

This part indicates the impact of the extrinsic factors

on the process execution. It illustrates the procedure

of the identiﬁcation and computation of these factors.

The procedure starts with computing the values of the

extrinsic variables from intertwining the information

of the intrinsic objects (both direct and indirect ob-

jects). This computation is based on the requirements

that are provided by the extrinsic factors. After this

computation, the relevant extrinsic factors could be

formed from these variables according to the context

needs. The deﬁned model can be applied to different

areas as it optimizes the execution of business proces-

ses based on different factors. In the next section, the

approach is applied to different scenarios which are

further discussed.

6 IMPLEMENTATION

This section discusses the ability to apply the metho-

dology of extrinsic dependencies to the running ex-

amples. These scenarios clarify the process depen-

dencies by representing the distinction between pro-

cess objects. In addition to that, they apply the proce-

dure of identifying and computing the impact of ex-

trinsic factors. As explained in the example of the

”repairing a car” process (cf. Sec. 2.1), one of the op-

timization issues occurs at the execution step of task

assignment if there is more than one eligible team.

This step could be optimized by identifying the most

eligible team. For this end, the direct and indirect ob-

jects are identiﬁed ﬁrst in Table 2.

Extrinsic Dependencies in Business Process Management Systems

637

Table 1: Extrinsic factor of ”repairing a car” process.

Context

requirement

Extrinsic

factor

Factor requirement

Factor va-

riable

Nearest team Accessibility

Location of the car

Distance

Team’s location

Table 2: Objects of the ”repairing a car” process.

Perspective

Direct

object

Property

Indirect

object

Property

Functional Repair a car

Data

Insurance

Digits

Car

speciﬁcation

Car model

Car’s

Location

Latitude

Longitude

Technical

issue

Issue

details

Organizational

Technical

team

Team name

Experience

High

Low

Availability

Available

Team size

Busy

Location

Latitude

Longitude

Operational

Technical

tools

Tools type

By applying the procedure of forming the extrin-

sic factors, a preference relation over all the available

teams could be identiﬁed. Generally, the ”accessibi-

lity” factor could be speciﬁed to optimize the execu-

tion by making the location-based assignment. The

formation of this factor is presented in Table 1.

As indicated in ”remotely conﬁguration of a de-

vice” process (cf. Sec. 2.2), one of the issues in the

execution steps is the simultaneous assignment of the

process to all (eligible) engineers regardless of their

context. Based on the approach here, the intrinsic ob-

jects of the process are clariﬁed in Table 4.

As explained in Sec. 2.2, the process execution

could be optimized based on two factors. The ﬁrst

one is the ”network security” which has an impact on

the authentication procedure followed. The second

one is the quality of the internet connection to avoid

any interruption might occur. The formation of these

factors are clariﬁed in Table 3.

To demonstrate the applicability of the introduced

approach for taking contextual factors relevant to a

process into account, an appropriate prototype is des-

cribed in this section. The prototype for our extrinsic

Table 3: Extrinsic factors of ”remotely conﬁguring a net-

work device” process.

Context

requirement

Extrinsic

factor

Factor requirement

Factor

variable

Reliable network

domain

Network

security

IP domain

Network

range

Highest QoS Connectivity Agents device properties QoS

dependencies methodology is based on web services.

It is developed as a web application which covers all

the steps related to deﬁning and executing the cor-

responding processes. It prioritizes those processes

which are related to assigning a task to participants.

The main components of the prototype are explained

as follows: Design core: It is used for initializing a

process in the system by simulating its fundamental

perspectives. Furthermore, it allows the identiﬁca-

tion of the relevant extrinsic factors and their requi-

rements. Execution core: It is responsible for exe-

cuting the deﬁned process. It is optimized with re-

spect to taking the impact of the identiﬁed extrinsic

factors into account. Therefore, it provides two exe-

cution types. The ﬁrst one is the normal execution,

which executes the process based on the information

of its fundamental perspectives. The second type is

the optimized execution, which takes the context in-

formation into account.

The application focuses on the geographical infor-

mation by using the ”repairing a car” example pro-

cess for the prototype application. Two agents named

agent1 and agent2 are used for running the experi-

ment. Both of them are logged into the system from

different locations. Both agents are notiﬁed by a nor-

mal execution procedure that they are agents recom-

mended, regardless of their contextual information.

As explained before, the accessibility factor could be

used to optimize process execution. This factor indi-

cates that the nearest agent to the location of the car is

preferred to perform the process. To this end, the pro-

totype fetches agent1 and agent2’s geographical loca-

tion by using web technologies, in particular, HTML5

Geolocation API (W3C, 2017), whereas the location

of the car is assumed when the process is deﬁned in

the system. Based on measured distance, agent1 is

identiﬁed as the nearest agent and notiﬁed that he is

recommended for the process, unlike agent2.

Table 4: Objects of ”Conﬁguring a network device”.

Perspective Direct object Property

Indirect

object

Property

Functional

Reconﬁguring

device remotely

Data

Ticket number

Device

Type

Organizational

Network

engineer

Name

Location

Latitude

Age Longitude

Gender

Used-device

IP domain

Job title type

Operational Application

ICEIS 2018 - 20th International Conference on Enterprise Information Systems

638

7 CONCLUSION

In this paper, a methodology for conceptualizing the

contextual factors relevant to a process was presen-

ted. The core of this methodology was the distinction

among process dependencies into intrinsic dependen-

cies and extrinsic dependencies. This distinction pro-

ved itself to facilitate modelling business processes

adaptable to context changes. The methodology clas-

siﬁed the context factors according to their relation

to the process. The main step of the methodology

was to deﬁne the approach of measuring the impact of

contextual factors that are related to the process des-

cription. It was shown that the entanglement between

process objects (direct and indirect) forms these va-

riables and computes its values. The process depen-

dencies model was introduced to present the concept

of contextualizing business processes. The applicabi-

lity of this approach was demonstrated by using the

prototype to test some scenarios with real informa-

tion. This presented the opportunity of using deve-

loped technologies for capturing the process context.

The results showed that the proposed approach leads

to improved execution of business processes and the

process assignment step in particular.

REFERENCES

Burkhart, T. and Loos, P. (2010). P.: Flexible business

processes - evaluation of current approaches. In

In: Proceedings Multikonferenz Wirtschaftsinformatik

(MKWI-2010).

de la Vara, Jose Luis, Ali, R., Dalpiaz, F., S

anchez, J., and

Giorgini, P. (2010). Business processes contextuali-

sation via context analysis. In Parsons, J., Saeki, M.,

Shoval, P., Woo, C., and Wand, Y., editors, Concep-

tual Modeling – ER 2010: 29th International Confe-

rence on Conceptual Modeling, Vancouver, BC, Ca-

nada, November 1-4, 2010. Proceedings, pages 471–

476. Springer Berlin Heidelberg, Berlin, Heidelberg.

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

editors (2013). Fundamentals of Business Process

Management. Springer Berlin Heidelberg and Im-

print: Springer, Berlin, Heidelberg.

Hammer, M. (2010). What is business process manage-

ment? In Handbook on Business Process Manage-

ment 1, pages 3–16. Springer-Verlag Berlin Heidel-

berg, Berlin, Heidelberg.

Imagery, N. and Agency, M. (2000). Department of defense

world geodetic system 1984: its deﬁnition and relati-

onships with local geodetic systems.

Jablonski, S. and Bussler, C. (1996). Workﬂow manage-

ment: Modeling concepts, architecture and implemen-

tation. ITP, Internat. Thomson Computer Press, Lon-

don [u.a.].

Modafferi, S., Benatallah, B., Casati, F., and Pernici, B.

(2005). A methodology for designing and managing

context-aware workﬂows. In Krogstie, J., Kautz, K.,

and Allen, D., editors, Mobile Information Systems

II: IFIP International Working Conference on Mobile

Information Systems, (MOBIS) Leeds, UK, December

6–7, 2005, pages 91–106. Springer US, Boston, MA.

Regev, G. and Wegmann, A. (2005). A regulation-based

view on business process and supporting system ﬂex-

ibility. In Faculdade de Engenharia da Universidade

do Porto, editor, Proceedings of the CAiSE, Modeling,

Development, and Support (BPMDS), pages 91–98,

Porto, Portugal.

Rosemann, M. and Recker, J. C. (2006). Context-aware pro-

cess design: Exploring the extrinsic drivers for pro-

cess ﬂexibility. In The 18th International Conference

on Advanced Information Systems Engineering. Pro-

ceedings of Workshops and Doctoral Consortium, pa-

ges 149–158.

Rosemann, M., R. J. and Flender, C. (2007). Contextuali-

zation of business processes. Int. J. Business Process

Integration and Management, Vol. 1.

Saidani, O. and Nurcan, S. (2007). Towards Context Aware

Business Process Modelling. In Workshop on Busi-

ness Process Modelling, Development, and Support,

page 1, Norway.

Sch

onig, S., Ackermann, L., and Jablonski, S. (2017). DPIL

Navigator 2.0: Multi-Perspective Declarative Process

Executio. In BPM Demos.

Sch

onig, S., Zeising, M., and Jablonski, S. (2014). Towards

location-aware declarative business process manage-

ment. In Abramowicz, W. and Kokkinaki, A., editors,

Business Information Systems Workshops: BIS 2014

International Workshops, Larnaca, Cyprus, May 22-

23, 2014, Revised Papers, pages 40–51. Springer In-

ternational Publishing, Cham.

Soffer, P. (2005). On the notion of ﬂexibility in business

processes. In Faculdade de Engenharia da Univer-

sidade do Porto, editor, Proceedings of the CAiSE,

volume 5 of Modeling, Development, and Support

(BPMDS), pages 35–42, Porto, Portugal.

Tao, Y., Zhu, G., Xu, Z., and Liu, B. (2008). A research on

bpm system based on process knowledge. In IEEE

Conference on Cybernetics and Intelligent Systems,

2008, Piscataway, NJ. IEEE.

W3C (2017). Geolocation api speciﬁcation.

Weske, M. (2012). Business Process Management: Con-

cepts, Languages, Architectures. Springer, Berlin and

Heidelberg, second edition edition.

Zhu, X., vanden Broucke, S., Zhu, G., Vanthienen, J., and

Baesens, B. (2016). Enabling ﬂexible location-aware

business process modeling and execution. Decision

Support Systems, 83:1–9.

Extrinsic Dependencies in Business Process Management Systems

639