An artifact centric view-based approach to modeling inter-organizational business processes

A. Bouguettaya, M. Hauswirth, and L. Liu (Eds.): WISE 2011, LNCS 6997, pp. 273–281, 2011.
© Springer-Verlag Berlin Heidelberg 2011
An Artifact-Centric View-Based Approach to Modeling
Inter-organizational Business Processes
Sira Yongchareon1
, Chengfei Liu1
, and Xiaohui Zhao2
1
Faculty of Information and Communication Technologies
Swinburne University of Technology, Victoria, Australia
{syongchareon,cliu}@swin.edu.au
2
Department of Computing, Faculty of Creative Industries and Business
Unitec Institute of Technology, Mount Albert, New Zealand
xzhao@unitec.ac.nz
Abstract. Over past several years, there have been increasing needs for more
efficient approaches to the design and implementation of inter-organizational
business processes. The process collaboration spanning organizational
boundaries is deemed to keep the organization autonomy, which means each
organization owns its freedom of modifying internal operations to meet their
private goals while satisfying the mutual objective with its partners. To achieve
these, we propose an artifact-centric view-based framework for inter-
organizational process modeling consisting of an artifact-centric collaboration
model, and a conformance mechanism between public view and private view to
support the participating organization’s customization and changes of their
internal operations while ensuring the correctness of collaboration process.
1 Introduction
Over past years, service-oriented architecture (SOA) has been serving an increasing
demand of IT for businesses to meet the challenges of the ever-changing market [7].
SOA particularly enables the business collaboration across organizations by the
means of web service composition to achieve the mutual goal of collaboration as well
as the individual goal of each participant. Service choreography is used to specify the
interaction between business parties whereby providing the agreed behavioral contract
between collaboration participants. Although service choreography technology
intends to provide a global view of the collaboration and allow each party to design
and implement its own portion, the existing choreography modeling approaches and
languages mainly describe the collaboration from the procedural perspective, and
focus on control-flow, message sequencing, etc. With this limited focus, the current
service choreography approaches cannot well support the three major requirements of
the collaboration: compliance, flexibility, and autonomy. The compliance refers to the
expectation that all parties must provide the services as they have agreed in the
contract. The flexibility means that each party has the freedom to change and
implement its own part in the collaboration while remaining autonomous.
In recent years, a new approach for modeling business processes has emerged,
namely artifact-centric process modeling, providing a higher level of robustness and
flexibility for describing process specification than traditional process-centric

274 S. Yongchareon, C. Liu, and X. Zhao
approaches [1, 2, 3, 4, 12]. In the process coordination, we observed that at a
particular state a message exchanged between organizations reflects the attribute
changes of the artifacts used in an organization’s internal process. Some states of the
artifacts of each individual party link with the global constraints in the collaboration.
Based on this observation and the three requirements, we propose a new paradigm of
inter-organizational process modeling namely artifact-centric collaboration model
together with the notion of conformance between public and private views to tackle
the compliance, autonomy, and flexibility issues of business collaboration.
The remainder of this paper is organized as follows. Section 2 introduces an
artifact-centric approach for modeling inter-organizational business processes.
Section 3 discusses the construction of public and private views. Section 4 reviews
the related works. Finally, the conclusion and future work are given in Section 5.
2 Modeling Inter-organizational Business Processes
In this section, we take supply chain collaboration as an example to illustrate and
motivate the artifact-centric approach to modeling inter-organizational business
collaboration, as shown Fig. 1. We initiate the discussion of this example by
identifying the types of involved business artifacts and describing how they are
modeled for this business collaboration. Here, we classify artifacts into two types: (1)
local artifact and (2) shared artifact. A local artifact is used internally within one
organization and can be used for supporting the coordination between intra-
organizational processes and inter-organizational processes. A shared artifact serves
as a contract between involved organizations, and it is used to indicate the agreed
business stages towards the completion of the collaborative process. Note that in the
implementation, shared artifacts act as messages sent and received between the
organizations. In Fig. 1 (a), the Supplier will ship the order only if the state of the
Purchase Order (PO) is in the confirmed state and the designated logistic company is
arranged by using the Shipping List (SL) and Shipping Order (SO) artifacts. We draw
the interrelation between artifacts as synchronization dependency (dashed-line).
Fig. 1. Overall collaboration and its public view of supply chain business processes
In the view-based approach to modeling inter-organizational processes, e.g.,
[7, 10, 11], a public view has to be constructed at the first stage based on the
integration of individual local process of each organization in the collaboration. For

An Artifact-Centric View-Based Approach 275
our artifact-centric model, only shared artifacts and their necessary (goal) states that
are required to be globally visible to other parties appear in the public view. After the
public view is built, the involved organizations should also have freedom to change
their own parts (i.e., their responsible parts of shared artifacts and their local artifacts)
later. As such, it is important to guarantee that such changes in the local process, or
private view, do not influence the correctness of the overall collaboration. We propose
the notion of view conformance as the validation mechanism of such changes in the
private view w.r.t. the public view. Due to page limitation, we do not detail how
public view is constructed in this paper; however, it can be achieved by adapting
artifact-centric process abstraction presented in our previous work [4].
Now, we define Artifact-Centric Collaboration Model (ACC model) extended
from the artifact-centric process model (ACP model) [3, 4, 16] to capture inter-
organizational processes. It consists of four core constructs: roles, artifacts, services,
and business rules. An artifact is either a local-typed or a shared-typed business entity
or an object involved in inter-organizational processes. An artifact class C is a tuple
(A, S, ) where A is set of a name-value pair attribute of scalar-typed value or nested
attribute structure (array-typed), S is a set of states, and ⊆ S{ } is a set of final
states where denotes initial state. A service is a task owned by one organization
in the collaboration and is used to perform (read/write) operations on artifact(s). A
business rule is defined by a condition-action style, which describes what condition a
service is invoked, and what (post) conditions attributes and state of artifacts must
satisfy after the invocation. Business rule r is triple (λ, β, v) where λ and β are pre-
condition and post-condition, respectively, containing two types of propositions: state
proposition and attribute proposition; and v is a service to be performed. A business
rule can be used to synchronize two or more artifacts, called synchronization
dependency. We denote D for its set, e.g., in Fig. 1, D(PO, SO) = { . } and D(PO,
SL) = { . , . }. Note that the synchronization dependency is transitive.
Definition 1: (Artifact-Centric Collaboration Model or ACC model). Let Π
denote an ACC model, and it is tuple (Z, V, R, L, ZL, VL) where,
− Z is a set of artifact classes, V is a set of services, and R is a set of business rules
− L is a set of organization roles participating in the collaboration
− ZL⊆ Z × 2L
is a set of artifact-role relations between artifacts and their
corresponding set of organization roles. Relation ( , { , ,…, })∈ZL means
that organization roles , ,…, involve in the changes of states of artifact
− VL⊆ V × L is a set of service-role relations between services and their
corresponding organization roles. Relation ( , )∈VL means that service is
provided by organization role .
Given ACC model Π, a lifecycle model of an artifact can be generated by deriving
corresponding business rules that are used to induce state transitions of such artifact.
A lifecycle of artifact class , denoted as , can be defined as tuple ( , T), where
T⊆ .S × Π.R × ( .S ∪ . ) is a 3-ary transition relation. T* is a reflexive
transitive closure of T. Next, we define the lifecycle occurrence (L-occurrence) for a
particular sequence of states occurring in the lifecycle. Based on L-occurrences, we
define a well-formed behavior property of an individual lifecycle in which is used

later for the discussion of a soundness property of the ACC model. A L-occurrence of
lifecycle is denoted as σ = ( , .., ) such that for every state s in σ, there
exists final state ∈ . and can be reached from through s by a
particular firing sequence of some business rules in R. We also denote = {σ , σ ,
.., σ } for a pairwise disjoint set of all possible L-occurrences in . Now, given
ACC model Π, Lifecycle of ∈ Π.Z is well-formed iff (1) there exists business
rule r∈Π.R such that r induce one and only one transition in ; and (2) for every
non-final state s∈ , there exists s in some L-occurrence of ; and (3) for every
final state ∈ , there exists L-occurrence σ ∈ such that is a last state of the
sequence in σ.
Definition 2: (Public and Private views). Given ACC model Π, a public view of Π,
denoted as p( Π , is the abstraction of Π , such that it represents only agreed
(abstracted) lifecycles of shared artifacts in Π. Private view Π is the local process of
Π for organization role l∈Π.L such that it represents only lifecycles of its own local
artifacts and the lifecycles of shared artifacts in p(Π .
As already discussed, when an agreed public view is constructed, involved
organizations can have their own freedom to modify their own private view. Here, we
define two behavioral-based modification operations, namely lifecycle modification:
(1) refine (responsible parts of) the lifecycle of shared artifacts, and (2) extend shared
artifacts with local artifact(s) that are required to coordinate with such shared artifacts.
Note that refining existing local artifacts is also considered as extending them to the
public view. To refine a shared artifact in a private view, we define lifecycle fragment
representing a partial lifecycle that is embedded into its existing lifecycle.
Definition 3: (Lifecycle Fragment or L-fragment). Given ACC model Π, artifact
∈Π. , and a set of states S = .S∪ . , we denote for a L-fragment which is
a nonempty connected sub-lifecycle of lifecycle . Let be a 4-tuple (S′, T′, Tin,
Tout) where S′ ⊆ .S{ }, T′ ⊆ S′ ×Π. × S′ ⊆ .T, a set of entry transitions Tin
= .T ∩ ((SS′) × Π. × S′)), and a set of exit transitions Tout = .T ∩ (S′ × Π.
× (SS′)) such that for every state s S′, there exist s in L-occurrences of .
Next, we define lifecycle modification in a private view for the refinement (by L-
fragments) of shared artifact and the extension of local artifacts.
Definition 4: (Lifecycle modification). Given public view p(Π , private view Π can
be modified based on p(Π with a lifecycle modification, denoted as Σ , and it is tuple
( , ), where is a set of extended lifecycles (of local artifacts) and is a
set of L-fragments }, = { , , ..., }, where (1<n<k)
is a L-fragment of ∈Π . refining the shared lifecycle of its in p(Π .
Fig. 2 shows an example of public and a private view of organization in
the collaboration where private view Π modify its public view with a lifecycle
modification consisting of extended local lifecycles of artifacts { , , }, refined L-
fragments { , } on artifact , and L-fragment on artifact . The

Fig. 2. Public view and its private view with lifecycle modification
organization also adds or modifies business rules used to synchronize local artifacts
with shared artifacts (within the refined L-fragments) to complete the overall process.
3 Validating Changes to Private View of a Collaboration Model
In this section, we discuss about the validation of changes made to private views
caused by lifecycle modification, i.e., how to ensure that the local changes to the
private view of one organization do not interfere with the whole collaboration. Note
that we confine our discussion only to the behavioural aspect of collaboration not the
validation of artifacts’ data. However, formal approaches to data verification can be
found in [3]. Here, we use state transition system to describe the behavioral aspect of
the ACC model, which can be generated by a composition technique presented in [6].
A state transition system of ACC model Π, namely ACC machine and denoted as
, is tuple (S, , T, ), where S and are a set of ACC states and the initial
state, respectively. T ⊆ S∪{ } × Π.R × G × S is a set of transitions and G is a set
of guards (state propositions). ⊆ S is a set of final states such that for every state s
in , s contains a combination of final state of every artifact in Π. is generated
by iteratively composing a lifecycle of each artifact in ∏.
Definition 5: (Soundness). Given ACC model Π and its ACC machine , Π is
sound iff (1) for every artifact ∈ Π.Z, the lifecycle of is well-formed; and (2) for
every artifact ∈ Π.Z, there exists a synchronization dependency between and
some other artifact; and (3) is well-formed and for every transition t∈ .T, guard
g of t contains no state referencing of any other artifact.
Condition (2) of the soundness property guarantees that the ACC model will consist
of only desired artifacts that are used in the collaboration, and the condition (3)
ensures the reachability of goal states of the collaboration and it also implies that a
domain of artifacts and their states referenced in the model is bounded. Next, we
define the view conformance between public and private views by checking whether
the lifecycle of the later covers the former. We generalize lifecycle coverage for both
artifact and ACC machine by reusing the L-occurrences definition of an individual
artifact for a machine. Given two machines = ( , , , ) and
CA
S1
S2
S3
L1.r1
L2.r3 L1.r5 S4
L1.r4L1.r2
C A
S1
S2
S3
L2.r3
S4
L1.r 4
L1.r 2
L1.rb
local-C1
S1 S2
local-C2
S1 S3S2
CB
S1 S2
L1.r6
S3
L1.r4L2.r7
init
C B
S1 S2 S3
L1.r 4L2.r7
Public view of collaboration Private view for org L1
Lf1
Lf2
L1.rd
L1.ra L1.rc
L1.rf
L1.re
d1
d2 d3
L1.rx
d6
L1.ry
L1.rz
Lf3
local-C3
S1 S2
d1
d3
d5
d6

= ( , , , ) and a set of states that exist in both .S and .S,
= ∩ , L-occurrences of is covered by L-occurrences of iff
, ∈ , ( , r, ) ∈ , ∈ , ∧ . Fig. 3 shows that
lifecycle (a) is not covered by (b) as in the occurrences of (b) state a can reach state c
without passing state b; in contrast, lifecycle (a) is covered by (c).
Fig. 3. Example of different refinement of L-fragments
Definition 6: (View conformance). Given private view Π modified based on public
view Π), Π conforms to Π) iff L-occurrences of the ACC machine of Π) is
covered by L-occurrences of the ACC machine of Π .
Although the view conformance is defined based on the overall behavior of the model,
it is important to discuss how we can validate the local modification. If such
modification is valid and correctly applied, then we can guarantee that it will preserve
the consistent behavior between the public and private views. Specifically, we apply
the lifecycle coverage checking between public and private views of a shared artifact
refined by L-fragments, and then we validate the whole modification with the
extended local artifacts in the private view. Now, given private view Π modified
based on sound public view Π) with lifecycle modification Σ , Σ is said to be valid
iff Σ can make Π sound and Π conforms to Π). Let the lifecycle of in
private view Π refine its public lifecycle of corresponding artifact in public view
Π ) with a set of L-fragments . L-occurrences of is covered by L-
occurrences of iff, for every ∈ , has a single entry state and a single
exit state. We name this as SESE L-fragment. For example, in Fig. 3, lifecycle (c)
consists of SESE L-fragments and refining lifecycle (a); thus, lifecycle (c)
covers lifecycle (a).
Next, we discuss how the synchronization is taken into account for the validation
of private views. First, we define the locally-bound property of the lifecycle of local
artifact that synchronizes with a shared artifact. Then, we use this property to induce
the coverage of all related artifacts in the private view. Let private view Π be
modified based on sound public view Π). For any local artifact ∈Π . and any
shared artifact ∈Π . Π).Z which is refined by a set of L-fragments , if
every synchronization dependency between lifecycle of and lifecycle of
occurs within one and only one fragment ∈ such that is SESE L-
fragment, then is locally-bound and preserves the soundness of Π . The
locally-bound property is transitive from one to another local artifact if a locally-
bound artifact synchronizes with another local artifact such that they do not
synchronize with any shared artifact, then such artifact is transitively locally-bound.

The transitivity is invalid if such artifact synchronizes with a non locally-bound
artifact. For example in Fig. 2, lifecycle of is locally-bound since synchronizations
and occur within SESE fragment of , and the lifecycle of is
transitively locally-bound via . In contrast, with , we can see that
synchronizations and occur in non-SESE L-fragment of artifact (even
though synchronizes with SESE L-fragment ); thus, the lifecycle of is not
locally-bound. Now, if organizations want to modify their private views using
lifecycle modification based on their public view, they need to ensure that their local
modifications do not violate the view conformance. Here, we show the conditions that
restrict the valid modification. Lifecycle modification Σ = ( , ) is valid for Π
based on sound Π), if, for every L-fragment ∈ , is SESE L-fragment and
for every lifecycle ∈ , is locally-bound, then L-occurrences of ACC
machine of Π is covered by L-occurrences of ACC machine of Π) (i.e., Π
conforms to Π)) and Π is sound. As the result, we can assert the soundness
property with the conformance by only checking the local lifecycle modification.
Whatever changes made on a private view will not impact on the behavior of
collaboration if they preserve valid modification.
4 Related Work and Discussion
In the area of artifact-centric process modeling, Nigam and Caswell [2] introduced the
concept of modelling business artifacts with lifecycles. Bhattacharya et al. [3] used
services to model activities and a set of business rules to capture and represent a
process model with the study of necessary properties such as reachability of goal
states, absence of deadlocks, and redundancy of data. Fritz et al. [12] studied
problems of goal-directed artifact-centric workflow construction. Hull et al. [1]
proposed an approach to interoperation of organizations hubs based on business
artifacts providing a centralized and computerized rendezvous point. Kuster et al. [8]
presented a notion of compliance of a business process model with object lifecycles
and a technique for generating the model from such set of lifecycles. In [4], we
proposed an artifact-centric process view framework to allow role-based process
abstraction that can be used to construct public views. All the above works serve as a
basis for this work. In the area of object-oriented design, Schrefl and Stumptner [5]
studied the consistency criteria of the inheritance of object life cycles. While their
work focused on the inheritance of single object, we adopt and extend their study of
consistent refinement to multiple lifecycles where the multiple inheritances of
artifacts may interact with each other in the collaboration.
In the area of cross-organizational workflow management, several approaches
presented in [10, 11, 13, 14, 15] were proposed to define, construct, and validate
public and private process views based on consistency rules and correctness-
preserving constraints. Van der Aalst et al. [7] proposed a process-oriented contract
agreed in the public view with a criterion for accordance between public and private
views. Their work is done based on process-centric setting while our work is on
artifact-centric setting. Lohmann and Wolf [9] presented an approach for artifact-
centric choreographies with the concept of agents and location-aware artifacts using
Petri-net model. They proposed a mechanism for automatic generation of an

interaction model that serves as a contract between the agents ensuring that specified
global goal states on the involved artifacts can be reached. Compared to our work,
their work only focused on the interaction model and did not consider the model of
local processes, while our approach presents the artifact-centric collaboration model
together with the validation mechanism for local model modification where
preserving global correctness.
5 Conclusions
This paper presents an artifact-centric view-based approach to modeling inter-
organizational business processes. We define the ACC model with the notion of
conformance of public and private views to address the choreography requirements.
The view conformance checking serves as the main resolving mechanism. We also
show that a modification on the private view of local organization can be theoretically
validated against the soundness properties for both public and private views.
References
1. Hull, R., Narendra, N.C., Nigam, A.: Facilitating Workflow Interoperation Using Artifact-
Centric Hubs. In: Baresi, L., Chi, C.-H., Suzuki, J. (eds.) ICSOC-ServiceWave 2009.
LNCS, vol. 5900, pp. 1–18. Springer, Heidelberg (2009)
2. Nigam, A., Caswell, N.S.: Business artifacts: An approach to operational specification.
IBM Systems Journal 42(3), 428–445 (2003)
3. Bhattacharya, K., Gerede, C., Hull, R.: Towards Formal Analysis of Artifact-Centric
Business Process Models. In: Alonso, G., Dadam, P., Rosemann, M. (eds.) BPM 2007.
4. Yongchareon, S., Liu, C.: Process View Framework for Artifact-Centric Business
Processes. In: Meersman, R., Dillon, T.S., Herrero, P. (eds.) OTM 2010. LNCS, vol. 6426,
pp. 26–43. Springer, Heidelberg (2010)
5. Schrefl, M., Stumptner, M.: Behavior-Consistent Specialization of Object Life Cycles.
ACM TOSEM 11(1), 92–148 (2002)
6. Lind-Nielsen, J., Andersen, H., Hulgaard, H., Behrmann, G., Kristoffersen, K., Larsen, K.:
Verification of Large State/Event Systems Using Compositionality and Dependency
Analysis. Formal Methods in System Design 18(1), 5–23 (2001)
7. Van Der Aalst, W.M.P., Lohmann, N., Masuthe, P., Stahl, C., Wolf, K.: Multipart
Contrats: Agreeing and Implementing Interorganizational Processes. The Computer
Journal 53(1), 90–106 (2010)
8. Küster, J.M., Ryndina, K., Gall, H.C.: Generation of Business Process Models for Object
Life Cycle Compliance. In: Alonso, G., Dadam, P., Rosemann, M. (eds.) BPM 2007.
9. Lohmann, N., Wolf, K.: Artifact-centric Choreographies. In: Maglio, P.P., Weske, M.,
Yang, J., Fantinato, M. (eds.) ICSOC 2010. LNCS, vol. 6470, pp. 32–46. Springer,
Heidelberg (2010)
10. Chebbi, I., Dustdar, S., Tata, S.: The view-based approach to dynamic inter-organizational
workflow cooperation. DKE 56, 139–173 (2006)

11. Van Der Aalst, W.M.P., Lohmann, N., Massuthe, P., Stahl, C., Wolf, K.: From Public
Views to Private Views – Correctness-by-Design for Services. In: Dumas, M., Heckel, R.
(eds.) WS-FM 2007. LNCS, vol. 4937, pp. 139–153. Springer, Heidelberg (2008)
12. Fritz, C., Hull, R., Su, J.: Automatic construction of simple artifact-based business
processes. In: ICDT 2009, vol. 361, pp. 225–238. ACM, New York (2009)
13. Zhao, X., Liu, C., Sadiq, W., Kowalkiewicz, M., Yongchareon, S.: Implementing Process
Views in the Web Service Environment. WWWJ 14(1), 27–52 (2011)
14. Zhao, X., Liu, C., Sadiq, W., Kowalkiewicz, M.: Process View Derivation and
Composition in a Dynamic Collaboration Environment. In: Chung, S. (ed.) OTM 2008,
Part I. LNCS, vol. 5331, pp. 82–99. Springer, Heidelberg (2008)
15. Zhao, X., Liu, C., Yang, Y., Sadiq, W.: Aligning Collaborative Business Processes - An
Organisation-oriented Perspective. IEEE TSMC 39(6), 1152–1164 (2009)
16. Yongchareon, S., Liu, C., Zhao, X., Xu, J.: An Artifact-centric Approach to Generating
Web-based Business Process Driven User Interfaces. In: Chen, L., Triantafillou, P., Suel,
T. (eds.) WISE 2010. LNCS, vol. 6488, pp. 419–427. Springer, Heidelberg (2010)

An artifact centric view-based approach to modeling inter-organizational business processes

More Related Content

Similar to An artifact centric view-based approach to modeling inter-organizational business processes (20)

Recently uploaded (20)

An artifact centric view-based approach to modeling inter-organizational business processes