Possibilistic Interorganizational Workflow Net for the Recovery Problem
Concerning Communication Failures
Leiliane Pereira de Rezende
1
, St
´
ephane Julia
1
and Janette Cardoso
2
1
Faculdade de Computac¸
˜
ao, Universidade Federal de Uberl
ˆ
andia, Av. Jo
˜
ao Naves de
´
Avila - 2160, Uberl
ˆ
andia/MG, Brazil
2
Institut Sup
´
erieur de l’A
´
eronautique et de l’Espace - ISAE, Av.
´
Edouard Belin - 10, Toulouse, France
Keywords:
Interorganizational Workflow Net, Possibilistic Petri Net, Deviations, Communication Failures, Process
Monitoring.
Abstract:
In this paper, an approach based on interorganizational WorkFlow nets and on possibilistic Petri nets is pro-
posed to deal with communication failures in business processes. Routing patterns and communication pro-
tocols existing in business processes are modeled by interorganizational WorkFlow nets. Possibilistic Petri
nets with uncertainty on the marking and on the transition firing are considered to express in a more realistic
way the uncertainty attached to communication failures. Combining both formalisms, a kind of possibilistic
interorganizational WorkFlow net is obtained. An example of communication failure at a process monitoring
level that precedes the presentation of a paper at a conference is presented.
1 INTRODUCTION
Workflow Management Systems are a key technology
for improving the effectiveness and efficiency of busi-
ness processes within an organization (van der Aalst,
1998b). Business processes represent the sequences
of activities that have to be executed within an organi-
zation to treat specific cases and to reach well defined
goals (Aalst and Hee, 2004). Over the last few years,
Business Process Management has become important
in order to raise service quality and a company’s per-
formance (Hofstede et al., 2010).
An organization produces value for its customers
by executing various business processes. Due to com-
plexity and variety of business processes, contempo-
rary organizations use information technology to sup-
port activities and possibly also automate their pro-
cesses. Business Process Management systems are
software systems used for automation of business pro-
cesses (Pesic, 2008).
In addition, as modern organizations have to
cope with complex administrative processes, Work-
flow Management Systems have to deal with work-
flow processes shared among multiple organizations.
Each business partner has to define private workflow
processes that are connected to other workflow pro-
cesses belonging to the other partners of the same
organization (Silva et al., 2013). The interorganiza-
tional workflow model corresponds then to a finite
set of WorkFlow nets loosely coupled through asyn-
chronous communication mechanisms (van der Aalst,
1998b).
Many papers have already considered Petri net
theory as an efficient tool for the modeling and analy-
sis of Workflow Management Systems (van der Aalst,
1998a; Aalst and Hee, 2004; Soares Passos and Julia,
2009). The WorkFlow nets, acyclic Petri net models
used to represent business processes, are defined in
(Aalst and Hee, 2004).
Soundness property is an important criterion
which needs to be satisfied when treating workflow
processes. In fact, good properties of well-defined
formal models such as WorkFlow nets can easily be
proved when business processes are following a rigid
structure that does not allow deviations from the pro-
cess description during real time execution.
However, recently, it was shown that business pro-
cesses do not easily map to a rigid modeling struc-
ture. The business processes are implemented such
that they “fit the system”, which can cause various
problems (Pesic, 2008). First, due to a mismatch be-
tween the preferred way of working and the system’s
way of working, companies may be forced to “run”
inappropriate business processes. Second, two paral-
lel realities may be created: the actual work is done
“outside the system” in one way , and later registered
in the system in another way (Pesic, 2008).
Attempts to consider a certain level of flexibility
432
Pereira de Rezende L., Julia S. and Cardoso J..
Possibilistic Interorganizational Workflow Net for the Recovery Problem Concerning Communication Failures.
DOI: 10.5220/0004863204320439
In Proceedings of the 16th International Conference on Enterprise Information Systems (ICEIS-2014), pages 432-439
ISBN: 978-989-758-027-7
Copyright
c
2014 SCITEPRESS (Science and Technology Publications, Lda.)
in process definition have already been proposed by
several authors.
In (Mohammed et al., 2007), a deviation-tolerant
approach in process execution is presented. In this ap-
proach, two process models coexist during the mon-
itoring of the process. The first one corresponds to
the expected behavior and the second is dynamically
built, based on the visible actions of human actors.
The two models of the process are permanently com-
pared and analyzed in order to detect deviations. Once
a deviation is detected, a deviation tolerance model at-
tached to the preset process is used to decide whether
to accept or to reject the deviation. The problem in
dealing with two models is that the monitoring activ-
ity can easily be overloaded implying a decrease in
the system’s performance.
In (Pesic et al., 2010), Aalst et al. created a declar-
ative workflow management system that uses con-
straint workflow models to achieve an optimal bal-
ance between flexibility and support. The basic idea
is that anything is allowed and possible unless explic-
itly forbidden. To implement this idea the authors
used a Linear Temporal Logic (LTL) and the ConDec
language. A limitation of this approach is the state
explosion that occurs when an automation process is
generated from the constraints for analysis.
Some authors, such as (van der Aalst et al., 2006;
Rozinat and van der Aalst, 2008; Munoz-Gama,
2010; van der Aalst et al., 2012), evaluated the con-
formance between the process model and the exe-
cution log of the process. For this, they developed
metrics for measuring the relationship between pre-
defined process models and actual results presented
in the form of event logs. The problem with this ap-
proach is that the verification is carried out after the
process execution.
A very promising alternative for dealing with flex-
ibility in business processes seems to be those ap-
proaches based on uncertain knowledge as that pre-
sented in (Cmpan and Oquendo, 2000). The model of
the process is then given through fuzzy sets and possi-
bilistic distributions that permit a natural representa-
tion of uncertain and imprecise information. This al-
ternative has already been shown in the flexible man-
ufacturing systems (Valette et al., 1982; Murata et al.,
1999; Asato et al., 2012).
One of the first studies which combines fuzzy and
possibilistic representation of information with the
precise structure of a Petri net when considering dis-
crete event systems is the one described in (Cardoso,
1999) and (Cardoso et al., 1999). The main feature
of possibilistic/fuzzy Petri nets is to allow one to rea-
son about the aspects of uncertainty and change in dy-
namic discrete event systems.
In (de Rezende et al., 2012) an approach based on
WorkFlow nets and on possibilistic Petri nets is pro-
posed to deal with non-conformance in Business Pro-
cesses. Its limitation is that it uses only a single pro-
cesses. In this paper, an approach based on interorga-
nizational WorkFlow nets and possibilistic Petri nets
is proposed to deal with communication failures in
business processes. In particular, a kind of possi-
bilistic interorganizational WorkFlow net will be de-
fined to treat the communication failures that occur
between the local WorkFlow nets.
The remainder of this paper is as follow: in sec-
tion 2, the definition of interorganizational WorkFlow
nets and soundness correctness criterion are provided.
In section 3, an overview of possibilistic Petri nets
is given. In section 4, the possibilistic interorgani-
zational WorkFlow net is presented and an example
based on a process that precedes the presentation of a
paper at a conference illustrates the approach. Finally,
section 5 concludes this work with a short summary,
an assessment based on the approach presented and
an outlook on future work proposals.
2 INTERORGANIZATIONAL
WORKFLOW NET
Before introducing the interorganizational WorkFlow
nets (IOWF-net) and the soundness property for these
nets, it is necessary to introduce the WorkFlow nets
(WF-nets) and soundness in the single organizational
context.
2.1 WorkFlow Net and Soundness
A Petri net that models a workflow process is called
a WorkFlow net (Aalst and Hee, 2004; van der Aalst,
1998a). A WorkFlow net satisfies the following prop-
erties (van der Aalst, 1998a):
• It has only one source place, named Start and only
one sink place, named End. These are special
places such that the place Start has only outgoing
arcs and the place End has only incoming arcs.
• A token in Start represents a case that needs to be
handled and a token in End represents a case that
has been handled.
• Every task t (transition) and condition p (place)
should be on a path from place Start to place End.
Soundness is a correctness criterion defined for
WorkFlow nets and is related to its dynamics. A
WorkFlow net is sound if, and only if, the follow-
ing three requirements are satisfied (Aalst and Hee,
2004):
PossibilisticInterorganizationalWorkflowNetfortheRecoveryProblemConcerningCommunicationFailures
433
• For each token put in the place Start, one and only
one token appears in the place End.
• When the token appears in the place End, all the
other places are empty for this case.
• For each transition (task), it is possible to move
from the initial state to a state in which that tran-
sition is enabled, i.e. there are no dead transitions.
2.2 Interorganizational WorkFlow Net
and Soundness
An interorganizational WorkFlow net (IOWF-net) is
essentially a set of loosely coupled workflow pro-
cesses modeled by a Petri net. Typically, there are n
business partners which are involved in one “global”
workflow process (Aalst, 1999). Each of the partners
has its own “local” workflow process, that is private,
and where a full control exists over it.
The local workflows interact at certain points, ac-
cording to a communication structure. There are two
types of communication: asynchronous communica-
tion (corresponding to the exchange of messages be-
tween workflows) and synchronous communication
(which forces the local workflows to execute specific
tasks at the same time) (Prisecaru and Jucan, 2008).
Synchronous communication corresponds to the melt-
ing of a number of transitions (Aalst, 1999).
In this paper, the synchronous case is not consid-
ered, since we consider that each organization con-
trols its own process. Only asynchronous communi-
cation protocols will be considered. Definition 1 for-
malizes the concept of an IOWF-net.
Definition 1. (IOWF-net) (van der Aalst, 1998b) An
interorganizational WorkFlow net(IOWF-net) is a tu-
ple IOW F = (PN
1
, PN
2
, ..., PN
n
, P
AC
, AC), where:
1. n ∈ N is the number of local WorkFlow nets
(LWF-nets);
2. For each k ∈ {1, ..., n} : PN
k
is a WF-net with
source place i
k
and sink place o
k
;
3. For all k, l ∈ {1, ..., n} : i f k 6= l, then (P
k
∪ T
k
) ∩
(P
l
∪ T
l
) =
/
0;
4. T
∗
=
S
k∈{1,...,n}
T
k
, P
∗
=
S
k∈{1,...,n}
P
k
, F
∗
=
S
k∈{1,...,n}
F
k
(relations between the elements of
the LWF-nets);
5. P
AC
is the set of asynchronous communication el-
ements (communication places);
6. AC ⊆ P
AC
× P(T
∗
) × P(T
∗
) is the asynchronous
communication relation
1
.
1
P(T
∗
) is the set of all non-empty subsets of T
∗
Each asynchronous communication element cor-
responds to a place name in P
AC
. The relation AC
specifies a set of input transitions and a set of out-
put transitions for each asynchronous communication
element.
The workflow which precedes the presentation of
a paper at a conference, presented in (van der Aalst,
1998b), will be used to understand the definition of
IOWF-net showed above. “This workflow can be con-
sidered to be an interorganizational workflow with
two loosely coupled workflow processes: (1) the pro-
cess of an author preparing, submitting and revis-
ing a paper, and (2) the process of evaluating and
monitoring submissions by the program committee.
In this case, there are two ‘organizations’ involved
in the interorganizational workflow: the author (AU)
and the program committee (PC). The author sends a
draft version of the paper to the program committee.
The program committee acknowledges the receipt and
evaluates the submission. The paper is accepted or re-
jected by the program committee. In both cases the
author is notified. If the paper is rejected, the work-
flow terminates, otherwise the author can start prepar-
ing the final version. After completing the final ver-
sion, a copy is sent to the program committee and the
program committee acknowledges the receipt of the
final version. If the final version is not received by
the program committee before a specific due date, the
author is notified that the paper is too late. A paper
which is too late will not be published in the proceed-
ings”.
Figure 1 shows the IOWF-net that models the pro-
cess described above. This IOWF-net has two LWF-
nets: AU and PC. The LWF-nets AU, on the left, mod-
els the local workflow of the author. The one on the
right, the LWF-nets PC, models the workflow proce-
dure followed by the program committee.
An IOWN-net which is composed of a number of
sound local workflows may be subject to synchroniza-
tion errors. In addition, it is also possible to have an
interorganizational workflow which is globally sound
but not locally sound (van der Aalst, 1998b). To de-
fine a notion of soundness suitable for IOWNF-nets,
Aalst in (van der Aalst, 1998b) defined the unfolding
of an IOWF-net into a WF-net.
In the unfolded net, i.e. the U(IOWF-net), all the
local WF-nets are connected to each other by a start
transition t
i
and a termination transition t
o
. Moreover,
a global source place i and a global sink place o have
been added in order to respect the basic structure of
a simple WF-net. Asynchronous communication el-
ements are mapped into ordinary places(P
AC
). The
result of the unfolding is a new WF-net.
The soundness property definition for interorgani-
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
434
Figure 1: An interorganizational workflow.
zational workflows is given below:
Definition 2. Soundness An interorganizational
WorkFlow net (IOWF-net) is sound iff it is locally
sound and globally sound. IOWF-net is locally sound
iff each of its local WorkFlow nets PN
k
is sound.
IOWF-net is globally sound iff U (IOWF-net) is sound.
The interorganizational workflow net shown in
Figure 1 is locally and globally sound. Then, the
U(IOWF-net) satisfy the soundness property.
3 POSSIBILISTIC PETRI NET
Possibilistic Petri nets are derived from Object Petri
nets (Sibertin-Blanc, 2001). In particular, in the ap-
proach presented in (Cardoso, 1999), a possibilistic
Petri net is a model where a marked place corresponds
to a possible partial state, a transition to a possible
state change, and a firing sequence to a possible be-
havior. The main advantage in working with possi-
bilistic Petri nets is that it allows for the updating of
a system state at a supervisory level with ill-known
information without necessarily reaching inconsistent
states.
A possibilistic Petri net model associates a possi-
bility distribution Π
o
(p) to the location of an object o,
p being a place of the net, thus allowing a possibilistic
distribution to the model:
• A Precise Marking: each token is located in only
one place (well-known state).
• An Imprecise Marking: each token location has
a possibility distribution over a set of places. It
cannot be asserted that a token is in a given place,
but only that it is in a place among a given set of
places.
Π
o
(p) = 1 represents the fact that p is a possible
location of o, and Π
o
(p) = 0 expresses the certainty
that o is not present in place p. Formally, a marking
in a possibilistic Petri net is then a mapping:
M : O × P −→ {0, 1}
where O is a set of objects and P a set of places. If
M(o, p) = 1, there exists a possibility of having the
object o in place p. On the contrary, if M(o, p) = 0,
there exists no possibility of having o in p. A marking
M of the net allows one to represent:
• A Precise Marking: M(o, p) = 1 and ∀p
i
6=
p, M(o, p
i
) = 0.
• An Imprecise Marking: for example, if there ex-
ists a possibility at a certain time to have the same
object o in two different places, p
1
and p
2
, then
M(o, p
1
) = M(o, p
2
) = 1.
A possibilistic marking will correspond in practice
to knowledge concerning a situation at a given time.
In a possibilistic Petri net, the firing (certain or
uncertain) of a transition t is decomposed into two
steps:
• Beginning of a Firing: objects are put into out-
put places of t but are not removed from its input
places.
• End of a Firing: that can be a firing cancellation
(tokens are removed from the output places of t)
or a firing achievement (tokens are removed from
the input places of t).
A certain firing consists the beginning of a firing
and an immediate firing achievement. A pseudo-firing
that will increase the uncertainty of the marking can
be considered only as the beginning of a firing (there
is no information to confirm whether the normal event
associated with the transition has actually occurred or
not). To a certain extent, pseudo-firing is a way of
realizing forward deduction.
PossibilisticInterorganizationalWorkflowNetfortheRecoveryProblemConcerningCommunicationFailures
435
The interpretation of a possibilistic Petri net is de-
fined by attaching to each transition an authorization
function η
x
1
,...,x
n
defined as follows:
η
x
1
,...,x
n
: T −→ {False,Uncertain, True}
where x
1
, ..., x
n
are the variables associated with the
incoming arcs of transition t (when considering the
underlying Object Petri net).
If o
1
, ..., o
n
is a possible substitution to x
1
, ..., x
n
for firing t, then several situations can be considered:
• t is not enabled by the marking but the associated
interpretation is true; an inconsistent situation oc-
curs and a special treatment of the net is activated;
• t is enabled by a precise marking and the inter-
pretation is true; then a classical firing (with cer-
tainty) of an object Petri net occurs;
• t is enabled by a precise marking and the interpre-
tation is uncertain; then the transition is pseudo-
fired and the imprecision is increased;
• t is enabled by an uncertain marking; if the inter-
pretation is uncertain, t is pseudo-fired;
• t is enabled by an uncertain marking and the inter-
pretation is true: a recovery algorithm, presented
in (Cardoso et al., 1989), is called and a new com-
putation of the possibility distribution of the ob-
jects involved in the uncertain marking is realized
in order to go back to a certain marking.
Concepts about possibilistic Petri nets will be il-
lustrated through a practical example in the next sec-
tion.
4 POSSIBILISTIC
INTERORGANIZATIONAL
WORKFLOW NET
The transitions in a classic Petri net represents the ex-
ecution of activities and a process state change. In
particular, each event occurring during the execution
of the process will be associated with a transition as a
boolean variable. Such a variable will be essentially
seen as an external value corresponding to a message
received from an activity (or received from another
process or sent to an activity). Possibly, internal val-
ues depending on certain token attributes will enable
some transitions too.
As pointed out in the introduction, the difficulty
to model business processes completely, considering
the set of all existing alternatives, is almost impossible
due to its complexity. From this, some inconsistencies
can occur between the model of the process and the
real process execution. A classical inconsistency in
the interorganizational case will be a communication
failure (a lost message or a delayed message that did
not appear at the right moment).
As each local process of an interorganizational
WorkFlow net is modeled by a Petri net, it can be di-
rectly executed using a specialized inference mecha-
nism called “token player algorithm” that allows for
a simplified monitoring of the represented process
model. A classical token player algorithm, as the one
defined in (Cardoso and Valette, 1997), is only based
on normal expected events. If an unexpected event
occurs, the process stop or needs to be repaired man-
ually.
A model of the process based on the routing struc-
ture of WorkFlow nets, on the communications be-
tween the local workflow processes of an interorgani-
zational WorkFlow net and on uncertain marking and
firing of a possibilistic Petri net will then produce a
kind of possibilistic interorganizational WorkFlow net
that will be able to deal with certain deviations, such
as failure communication, unexpected, delayed or lost
events, in business process monitoring.
Figure 2: AU process using possibilistic WorkFlow net.
The process that precedes the presentation of a pa-
per at a conference, described in Subsection 2.2 and
represented in Figure 1, will be used to illustrate the
approach. In this paper, only the AU process will be
transformed into a possibilistic WorkFlow net as il-
lustrated in Figure 2 and the communication places
are considered as external events associated with the
transitions of the local WorkFlow net.
< p > is an object belonging to the class “Paper”,
x is a variable of the same class “Paper” and all places
of the model belong to the class “Paper” too. Each
transition has an interpretation and an action attached
to it defined by the designer. The interpretation is
used to manage the occurrence of each event in the
system by imposing restrictions on the firing of tran-
sitions. The action is an application that involves the
attributes or methods of formal variables associated
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
436
with incoming arcs allowing for the modification of
some specific attributes. Some actions can be exe-
cuted only after the certain firing of the transition and
others, with a modifier attached to it, for any type of
firing
Considering the AU process, represented by the
Figure 2, the expected behavior, after the author sends
a draft version of the paper to the program committee,
is to receive two messages sent by the program com-
mittee, one referring to the notification of receipt of
the draft and another concerning the acceptance or re-
jection of the paper.
A deviation of the expected behavior can easily
occur if, for example, the program committee does
not notify the receipt of the draft or delay the send-
ing of the notification of the acceptance or rejection
of the paper. Logically, the author should continue
preparing the article even without knowing if it was
accepted or rejected (firing of transition t
3
or t
4
).
Knowing that the object instances of class “Paper”
have the attribute date, responsible for defining the
time limit needed to wait for normal event occurrence
time, the interpretations of transitions t
2
, t
3
, t
4
, t
5
and
t
6
are given by the following distributions:
η
x
(t
2
) =
true i f (ack dra f t)
uncertain i f (τ ≥ x.date) ∧ (¬ack dra f t)
f alse otherwise
η
x
(t
3
) =
true i f (re ject)
uncertain i f (τ ≥ x.date) ∧ (¬re ject)
f alse otherwise
η
x
(t
4
) =
true i f (accept)
uncertain i f (τ ≥ x.date) ∧ (¬accept)
f alse otherwise
η
x
(t
5
) =
true i f (accept) ∧ (¬too late)
uncertain i f ¬ (accept ∨ too late)
f alse otherwise
η
x
(t
6
) =
true i f (too late)
f alse otherwise
where τ is the current time, ack dra f t is true when the
program committee acknowledges the receipt of the
draft, re ject is true when the paper is rejected by the
program committee, accept is true when the program
committee accept the paper and too late is true when
the paper is received after the deadline.
The normal expected behavior of the AU process,
after sending the draft to the program committee, cor-
responds to the recognition (ack dra f t) of the receipt
of the draft and after the acceptance (accept) or re-
jection (re ject) of the paper before a specific due date
indicated by the attribute date associated to the ob-
ject < p >. If the expected messages are received in
time indicated by the attribute date associated to the
corresponding object, all the transition firing will be
certain and all the markings will be precise.
An abnormal behavior will happen if the current
time reaches the value of the attribute date of the ob-
ject < p > and no message has been received from
the program committee for the corresponding event.
In this case some pseudo-firing will have to occur and
the imprecision about some objects will increase.
Let’s assume the transition t
1
is fired at date τ = 10
(Figure 3(a)) and that the actions associated to tran-
sitions t
1
, t
2
, t
4
and t
5
responsible for updating the
waiting time limit for the event’s occurrence are the
following:
Action(t
1
) : x.date = τ + 5
Action(t
2
) : x.date = τ + 45
Action(t
4
) : x.date = τ + 5
Action(t
5
) : x.date = τ + 45
if ack dra f t = f alse and re ject = f alse all the time
and accept = true at τ = 80, the following scenario
will occur:
• at current time τ = 15, the transition t
2
is en-
abled by a certain marking and its interpretation
is uncertain (η
<p>
(t
2
) = uncertain). Then, t
2
is
pseudo-fired executing the action associated to it
(Action(t
2
) : x.date = τ + 45 = 60) (Figure 3(b));
• at current time τ = 60, the transitions t
3
and t
4
are
enabled by an uncertain making and the interpre-
tation attached to them is uncertain (η
<p>
(t
3
) =
uncertain, η
<p>
(t
4
) = uncertain). Then, t
3
and t
4
are pseudo-fired executing just the action associ-
ated to transition t
4
(Action(t
4
) : x.date = τ + 5 =
65) (Figure 3(c));
• at current time τ = 65, the transition t
5
is en-
abled by an uncertain marking and its interpreta-
tion is uncertain (η
<p>
(t
5
) = uncertain). Then,
t
5
is pseudo-fired executing the action associated
to it (Action(t
5
) : x.date = τ + 45 = 110)(Figure
3(d));
• at current time τ = 80, the transition t
4
is enabled
by an uncertain marking and its interpretations be-
comes true (η
<p>
(t
4
) = true). This situation oc-
curs because the notification of the receipt of the
draft by the program committee never arrives and
the paper was accepted but with a delay in the re-
sponse on the part of PC process. Consequently a
recovery algorithm, presented in (Cardoso et al.,
1989), is called to go back to the certain marking
of Figure 3(e), archiving the pseudo-fired transi-
tion t
4
and canceling the pseudo-firing of transi-
tions t
3
and t
5
.
This scenario in a classical WorkFlow net would
lead to an inconsistency due to the interpretation asso-
ciated to transition t
4
becoming true without the pres-
PossibilisticInterorganizationalWorkflowNetfortheRecoveryProblemConcerningCommunicationFailures
437
(a) τ = 10 (b) τ = 15 (c) τ = 60 (d) τ = 65 (e) τ = 80
Figure 3: Simulation results of the scenario.
ence of a token in a
2
to enable the transition t
4
as il-
lustrated in the Figure 4 in highlight. In practice, a
human actor would easily deal with such a delay but
a classical WorkFlow net would reach a global incon-
sistency.
Figure 4: An possible inconsistency in a classical Work-
Flow net.
To take into account the kind of incident with an
ordinary Petri net based on the classical token player
defined in (Cardoso and Valette, 1997), several new
transitions should be created to consider all possi-
ble abnormal scenarios. As a consequence, the cor-
responding graph would rapidly become completely
unreadable and complex.
Each workflow process of the interorganizational
process structure will follow the behavior of the pos-
sibilistic token player algorithm given in figure 5. In
particular, such an inference mechanism will ensure
that small deviations within the interorganizational
structure will not necessarily lead to inconsistencies.
Figure 5: Possibilistic token player algorithm associated to
autonomous local processes.
5 CONCLUSIONS
In this article, a possibilistic interorganizational
WorkFlow net model was presented with the pur-
pose of dealing with communication failures in busi-
ness processes. Combining the routing structure of
a WorkFlow net, the communication mechanisms be-
tween the local workflow processes and the uncertain
reasoning of possibilistic Petri nets, it was possible
to deal with some communication failures that can
happen during the execution of an interorganizational
workflow process. Such an approach was applied to a
process that precedes the presentation of a paper at a
conference.
The occurrence of lost, delayed, spurious events
are handled by pseudo-fired until the moment a cor-
rect event occurs to return to a certain marking. Com-
paring the behavior of this approach with other works
dealing with the problem of deviations, its main ad-
vantage is that the deviations are discovered and re-
covered at the moment of execution (at the execu-
ICEIS2014-16thInternationalConferenceonEnterpriseInformationSystems
438
tion time) and several possibilities of the execution are
created through the uncertain reasoning. In addition,
the fact that a formal process model which allows
one to prove some of the good properties, like the
soundness property for example, was combined with
a possibilistic approach which is very well adapted to
the concept of flexibility and robustness in processes.
However, the approach does not permit models with
loops and pseudo-firings in transitions that have al-
ready been pseudo-fired.
As a future work proposal, it will be interesting to
present a communication failures recovery approach
in a interorganizational workflow process not neces-
sarily sound, knowing that in practice, the inherent
flexibility of legacy systems does not always allow
for the production of a process model that respects
the soundness property.
REFERENCES
Aalst, W. (1999). Interorganizational workflows: An ap-
proach based on message sequence charts and petri
nets. Systems Analysis - Modelling - Simulation,
34:335 – 367.
Aalst, W. v. d. and Hee, K. v. (2004). Workflow Manage-
ment: Models, Methods, and Systems. MIT press,
Cambridge, MA.
Asato, O. L., Dobrianskyj, G. M., Junqueira, F., dos San-
tos Filho, D. J., and Miyagi, P. E. (2012). Process con-
trol system considering the machines functional flexi-
bilities. In DoCEIS, pages 133 – 142.
Cardoso, J. (1999). Time fuzzy petri nets. In Fuzziness
in Petri Nets, volume 22, pages 115 – 145. Physica-
Verlag.
Cardoso, J. and Valette, R. (1997). Redes de Petri.
DAUSFC.
Cardoso, J., Valette, R., and Dubois, D. (1989). Petri nets
with uncertain markings. In Applications and Theory
of Petri Nets, pages 64 – 78.
Cardoso, J., Valette, R., and Dubois, D. (1999). Possibilistic
petri nets. Systems, Man, and Cybernetics, Part B:
Cybernetics, IEEE Transactions on, 29:573 –582.
Cmpan, S. and Oquendo, F. (2000). Dealing with software
process deviations using fuzzy logic based monitor-
ing. SIGAPP Appl. Comput. Rev., 8:3 – 13.
de Rezende, L. P., Julia, S., and Cardoso, J. (2012). Pos-
sibilistic workflow nets to deal with non-conformance
in process execution. In Systems, Man, and Cybernet-
ics (SMC), 2012 IEEE International Conference on,
pages 1219 – 1224.
Hofstede, A., Aalst, W., Adams, M., and Russell, N. (2010).
Modern Business Process Automation: YAWL and its
Support Environment. Springer-Verlag, Berlin.
Mohammed, K., Redouane, L., and Bernard, C. (2007). A
deviation-tolerant approach to software process evo-
lution. In Ninth international workshop on Princi-
ples of software evolution: in conjunction with the 6th
ESEC/FSE joint meeting, pages 75 – 78.
Munoz-Gama, J. (2010). Algorithms for process confor-
mance and process refinement. Master’s thesis, Uni-
versitat Polit
`
ecnica de Catalunya (UPC).
Murata, T., Suzuki, T., and Shatz, S. (1999). Fuzzy-Timing
High-Level Petri Nets (FTHNs) for Time-Critical Sys-
tems, volume 22, pages 88 – 114. Physica-Verlag.
Pesic, M. (2008). Constraint-based Workflow Management
Systems: Shifting Control to Users. Phd thesis, Eind-
hoven University of Technology.
Pesic, M., Schonenberg, H., and Aalst, W. (2010). Declar-
ative workflow. In Modern Business Process Automa-
tion: YAWL and its Support Environment, chapter VI,
pages 175 – 201. Springer-Verlag, Berlin.
Prisecaru, O. and Jucan, T. (2008). Interorganizational
workflow nets: a petri net based approach for mod-
elling and analyzing interorganizational workflows. In
Enterprise & Organizational Modeling and Simula-
tion, pages 64 – 78.
Rozinat, A. and van der Aalst, W. M. P. (2008). Confor-
mance checking of processes based on monitoring real
behavior. Information Systems, 33:64 – 95.
Sibertin-Blanc, C. (2001). Cooperative objects: Princi-
ples, use and implementation. In Concurrent Object-
Oriented Programming and Petri Nets, pages 216–
246.
Silva, L. d. F., Soares Passos, L. M., Soares, M. d. S., and
Julia, S. (2013). Siphon-based deadlock prevention
policy for interorganizational workflow net design. In
14th IEEE International Conference on Information
Reuse and Integration, pages 293 – 300.
Soares Passos, L. M. and Julia, S. (2009). Qualitative anal-
ysis of workflow nets using linear logic: Soundness
verification. In Systems, Man and Cybernetics, 2009.
SMC 2009. IEEE International Conference on, pages
2843 –2847.
Valette, R., Courvoisier, M., and Mayeux, D. (1982). Con-
trol of flexible production systems and petri nets. In
European Workshop on Applications and Theory of
Petri Nets, pages 264 – 277.
van der Aalst, W. M. P. (1998a). The application of petri
nets to workflow management. Journal of Circuits
Systems and Computers, 8:21 – 66.
van der Aalst, W. M. P. (1998b). Modeling and analyz-
ing interorganizational workflows. In Proceedings of
the 1998 International Conference on Application of
Concurrency to System Design (CSD’98), pages 262
– 272.
van der Aalst, W. M. P., Adriansyah, A., and van Dongen, B.
(2012). Replaying history on process models for con-
formance checking and performance analysis. Wiley
Interdisciplinary Reviews: Data Mining and Knowl-
edge Discovery, 2:182 – 192.
van der Aalst, W. M. P., Dumas, M., Ouyang, C., Rozi-
nat, A., and Verbeek, H. M. W. (2006). Choreography
conformance checking: An approach based on bpel
and petri nets. In The Role of Business Processes in
Service Oriented Architectures.
PossibilisticInterorganizationalWorkflowNetfortheRecoveryProblemConcerningCommunicationFailures
439
