TRANSPARENCY IN CITIZEN-CENTRIC SERVICES

A Traceability-based Approach on the Semantic Web

Ivo J. Garcia dos Santos and Edmundo R. Mauro Madeira

Institute of Computing, University of Campinas (UNICAMP), P.O. Box 6176, Campinas SP, Brazil

Keywords:

e-Government, Semantic Web Services, Traceability, Middleware.

Abstract:

The search for effective strategies to increase the transparency in public administration processes is becoming

a key issue for governments and for the success of the new citizen-centric services and applications. Also, the

application of service-oriented architectures, semantics and ontologies is gaining momentum as an alternative

to fulﬁll the inherent e-Government interoperability and dynamism demands. Considering the challenges

introduced by this new scenario, this paper contributes proposing an approach to monitor and audit composite

e-Government services. The solution is based on a set of Traceability Policies modeled and implemented over

a semantically-enriched and service-oriented middleware (CoGPlat).

1 INTRODUCTION

The demands for the creation of mechanisms to

increase the transparency of the public adminis-

tration processes have dramatically increased over

the recent years. It represents a requirement for

every democracy identiﬁed since the origins of the

modern republic: Thomas Jefferson once wrote that

”whenever the people are well-informed, they can

be trusted with their own government” (Jefferson,

1789). In parallel, new citizen-centric services and

applications are becoming more and more investi-

gated, mainly due to the proliferation of the new ICTs

(Information and Communication Technologies).

The use of Service-oriented architectures and, more

recently, semantics and ontologies, is also gaining

momentum to enable interoperability and to in-

crease the dynamism of e-Government applications.

Considering this scenario, this paper contributes

by presenting a strategy to achieve transparency in

composite e-Government Services based on a set

of Traceability Policies modeled and implemented

over a semantically-enriched and service-oriented

middleware (CoGPlat). The remainder of this paper

is organized as follows: Section 2 introduces the con-

cepts and technologies related to the contributions of

the paper and presents relevant related work; Section

3 introduces the proposed approach to increase trans-

parency in composite e-Government services; and

ﬁnally, Section 4 presents the concluding remarks.

2 LITERATURE REVIEW

The interoperability, deﬁned as ”the ability of two

or more systems or components to exchange infor-

mation and to use the information that has been ex-

changed” (IEEE, 1990), is a fundamental requirement

in the context of distributed and dynamic applica-

tions. In order to achieve higher levels of interop-

erability, many systems are being implemented fol-

lowing the Service-Oriented Architecture (SOA) ap-

proach (Papazoglou and Georgakopoulos, 2003) and

its most famous manifestation, the Web Services tech-

nologies. One of the current Web biggest limita-

tions is the lack of support to describe the semantics

of data. To overcome this weakness, the so-called

Semantic Web (Berners-Lee et al., 2001) proposes

a scenario where data becomes meaningful to ma-

chines and can be automatically processed and un-

derstood. In it, ontologies play a fundamental role in

the deﬁnition of the concepts that are used to annotate

data (Medjahed, 2004). Speciﬁcations such as RDF

(Resource Description Framework) and RDF-Schema

were among the ﬁrst W3C (World Wide Web Consor-

tium) initiatives to model meta-data related to Web re-

sources. Later, the OWL (Ontology Web Language)

extended RDF, augmenting its vocabulary with the in-

clusion, for instance, of new class relationships. It

is the current W3C standard for specifying ontolo-

gies on the Web. The possibility of applying a simi-

lar strategy to semantically describe services, opening

the road towards their automatic discovery, invocation

184

J. Garcia dos Santos I. and R. Mauro Madeira E. (2008).

TRANSPARENCY IN CITIZEN-CENTRIC SERVICES - A Traceability-based Approach on the Semantic Web.

In Proceedings of the Tenth International Conference on Enterprise Information Systems - SAIC, pages 184-189

DOI: 10.5220/0001720601840189

 SciTePress

and composition, motivated the proposal of different

approaches for the deﬁnition of what was baptized

as the Semantic Web Services. OWL-S (Semantic

Markup for Web Services), for instance, combines a

set of inter-related OWL ontologies that deﬁne terms

used in service-oriented applications. Besides OWL-

S (adopted in the work described in this paper) two

other proposals already play an important role in the

Semantic Web Services scenario: the WSMO (Web

Services Modeling Ontology) (W3C, 2005)) and the

SA-WSDL (Semantic Annotations for WSDL and

XML Schema) (W3C, 2007), recently adopted as a

W3C recommendation.

Techniques which enable markup and automated

reasoning technology to describe, simulate, test, and

verify compositions of Web services are discussed in

(Narayanan and McIlraith, 2002). The authors deﬁne

the semantics for a relevant subset of OWL-S in terms

of a ﬁrst-order logical language (Situation Calculus).

With the semantics in hand, service descriptions are

encoded in a Petri Net formalism. The implemented

system is able to read in OWL-S service descriptions

and perform simulation, enactment and analysis. The

use of advanced workﬂow and activity concepts in the

composition of Web services is the proposal of (Fileto

et al., 2003). The approach is called POESIA (Pro-

cesses for Open-Ended Systems for Information Anal-

ysis), an open environment for developing Web appli-

cations using metadata and ontologies to describe data

processing patterns developed by domain experts. It

supports Web service composition using domain on-

tologies with multiple dimensions (e.g., space, time,

and object description).

The electronic Government (e-Government) do-

main includes the ”set of all processes which serve

decision-making and services in politics, government

and administration and which use information and

communication technologies” (KBSt, 2006). Re-

cently a new approach to e-Government is gaining

momentum: the citizen-centric government, where

citizens and businesses are considered customers of

the public administration, so that their needs come

ﬁrst, rather than bureaucracy or other imperatives in-

side the government machine (GOV3, 2006). In this

context, usually a government-wide service-oriented

architecture is applied to develop a single place that

offers access to all government informational and

transactional services. Several research efforts in the

e-Government domain can be found in the literature.

For instance, a system which automatically generates

Web services customized to citizens’ needs and also

to government laws and regulations is presented in

(Medjahed and Bouguettaya, 2005). It proposes three

levels of service customization: the Citizen level, the

Service level and the User interface level. A meta-

data ontology, used to describe e-Government ser-

vices and operations, is also introduced. An approach

for the semi-automated design of data ﬂows between

Web Services that are semantically described using

different ontologies and data representations is intro-

duced in (Barnickel et al., 2006), including a rule-

based mechanism for user-transparent mediation be-

tween ontologies spanning multiple application do-

mains.

3 ENABLING TRANSPARENCY

As already mentioned, transparency is becoming a

fundamental requirement in citizen-oriented applica-

tions. One of the possible strategies to improve trans-

parency is to guarantee traceability at the service

level. Traceability is deﬁned by the International

Organization for Standardization (ISO, 1994) as the

”ability to trace the history, application or location

of an entity by means of recorded identiﬁcation”. We

introduce a strategy based on traceability policies to

regulate the execution monitoring of composite e-

Government services and detail it next.

3.1 The Middleware

The strategy presented in this paper is modeled and

implemented over an e-Government service middle-

ware called CoGPlat (Citizen-oriented e-Government

Platform) (Santos et al., 2005). Its main goal is to

support applications that enable the interaction and

collaboration among governmental entities, organiza-

tions and citizens and its infrastructure includes (see

Figure 1): a Service Bus, an interface between the

middleware services and the applications; four core

facilities (described next); a Service Discovery and

Execution Layer, responsible for selecting the Se-

mantic Web Services to participate in the compo-

sitions and also for interacting with those services

at process run-time; and a set of Support Services

which provides security, persistence, reliable messag-

ing and transaction support to the processes running

over the platform.

The four middleware core facilities are (see also

Figure 1): the Transparent Services Center, respon-

sible for dynamically building the service composi-

tions according to the application requests; the Meta-

model Management Center, which offers services

and tools to manage the models, metamodels and on-

tologies used in the description of services, composi-

tions, processes and entities; the E-Governance and

E-Democracy Center, which delivers generic ser-

TRANSPARENCY IN CITIZEN-CENTRIC SERVICES - A Traceability-based Approach on the Semantic Web

185

Figure 1: CoGPlat Architecture Overview.

vices which aim to facilitate the decision-making pro-

cesses and to increase citizen participation in the pub-

lic administration; and the Traceability and Audit-

ing Center, which offers services and tools to moni-

tor running processes and also to audit processes that

have already been concluded. The contributions of

this paper are concentrated in this last facility.

3.2 Policies

To effectively compose e-Government services, is-

sues like the autonomy of the entities, data privacy,

process traceability and an efﬁcient identity manage-

ment usually need to be considered. In order to enable

compositions which handle these demands, CoG-

Plat implements a set of Interaction Policies which

were ﬁrst proposed and successfully applied in an e-

Business scenario (Santos and Madeira, 2006) and

then extended to match the new requirements de-

manded by e-Government applications. These poli-

cies are applied to regulate the collaborations that

take place over the middleware infrastructure - we

consider that two entities collaborate when both par-

ticipate in the same composite service (as providers

and/or consumers) and there is at least one informa-

tion or message exchange between them through the

middleware facilities. The CoGPlat Interaction Poli-

cies are classiﬁed into the following categories (all

policy terms and relations are speciﬁed in an OWL

ontology):

1. Entity Autonomy Policies. determine the level

of control the platform (and therefore the applica-

tions running over it) may have over the internal

operations of a composite service;

2. Data Privacy Policies. determine the collabora-

tion levels on the interactions between two entities

that participate on the same composite service,

deﬁning how the privacy of the data exchanged

between them is handled;

3. Identity Management Policies. establish where

to provide mechanisms like anonymousness, iden-

tity theft protection and non-repudiation for citi-

zens and entities;

4. Service Traceability Policies. determine to

what extent the operations of a composite service

should be tracked by the middleware. This is the

category closely related to the contributions pre-

sented in this paper and will be further detailed

next.

An abstract composition and individual abstract activ-

ities can be semantically annotated with the following

Service Traceability policies:

• Strong-traceability. all possible events both at

composition and at activity/service level are mon-

itored. Customized traceability requirements can

also be speciﬁed (e.g.: monitor the behavior of

an speciﬁc property/value throughout the instance

execution);

• Weak-traceability. only composition level events

are monitored. No customizations are allowed;

• Zero-traceability. There is no traceability at all.

Indeed, if this policy is speciﬁed, there must be no

trace (neither during nor after) of a given compo-

sition instance.

3.3 The Composition Process

One of the main goals of CoGPlat is to facilitate the

development and operation of new e-Government ap-

plications that present higher levels of dynamism and

citizen-orientation. Given the quantity and diverse-

ness of public administration agencies, to integrate

and interoperate the services delivered by those agen-

cies become a challenging task. The service com-

position process described next tries to shift to the

middleware level the solution for those integration

problems. It is important to understand how CoGPlat

builds composite services before a deeper discussion

on the Traceability mechanisms is held. The compo-

sition process has the following stages (see Figure 2):

ICEIS 2008 - International Conference on Enterprise Information Systems

186

Figure 2: Composition Process in CoGPlat.

(1) An application sends a request (an OWL-S

proﬁle) to the Transparent Services Center for

a composite service. The request includes the

IOPEs (Input, Output, Precondition, Effects)

and the demanded interaction policies;

(2) The Metamodel Management Center performs

a search in the repository for an abstract com-

position that satisﬁes the application require-

mens. An OWL-S matchmaker is used in this

stage;

(3a, 4) If an abstract composition is found, a search

for concrete services that will execute the ab-

stract composition activities is started. The in-

teraction policies are, together with the IOPEs,

used as selection criteria. (3b, 6, 7a, 7b):

Else, a search for external services (not pre-

registered in the platform) is tried;

(5a) If a concrete composition has been success-

fully built, its speciﬁcation is prepared for ex-

ecution. (5b, 6, 7a, 7b): Else, an external

search is tried (as in (3b)).

The chances of success on the initial stages of the

composition process (1 and 2 in Figure 2) are directly

related to the amount and diversity of abstract compo-

sitions registered in the platform. These abstract com-

positions describe the generic public processes that

may run over the platform combining services from

different agencies. No concrete binding is necessary

in design time - actually only service classes, semanti-

cally annotated with IOPEs and Policies are included

in the abstract composition ﬂow.

3.4 Traceability and Auditing Center

The Traceability and Auditing Center implements

the necessary mechanisms to guarantee that the ac-

tive traceability policies are respected during the ex-

ecution of the composite services. Its internal in-

frastructure is composed of the following elements

(see also Figure 3): the Standard Tracking Service

which collects all data from the generated tracking

events, according to the active policy, and saves it

into a repository; the Real-time Tracking Service

which collects only representative data from the track-

ing events, also according to the active policy, but pro-

vides it in real-time to the platform client applications

instead of saving it to a repository; the Auditing Ser-

vice which offers mechanisms to access tracking data

of already concluded processes; and a Tracking Data

Repository that stores the information generated by

the tracking service events.

Figure 3: Traceability and Auditing Center Infrastructure.

3.5 Composition Example

In order to illustrate how the Traceability Policies are

evaluated and executed in CoGPlat, let’s follow the

example of a generic process of requesting a doc-

ument. The process is composed of the following

steps: (1) Validation of the requester personal infor-

mation; (2) Payment of administrative fees; (3) For-

ward of the request to the responsible government of-

ﬁce (determined by the document type).

The composition process follows the steps previ-

ously presented in Figure 2. First, the application

sends a request containing IOPEs and demanded poli-

cies. A search is then performed in the abstract com-

positions repository. In Figure 4 the selected abstract

composition is shown. Note that the composition and

the individual activities are annotated with IOPEs and

with the supported traceability policies.

TRANSPARENCY IN CITIZEN-CENTRIC SERVICES - A Traceability-based Approach on the Semantic Web

187

Figure 4: Abstract Composition with Policies and IOPEs.

Next, according to the process in Figure 2, the

semantic annotations are used to select concrete ser-

vices that will implement each of the activities.

When the composition starts to execute, the Stan-

dard Tracking Service of the Traceability and Au-

diting Center is activated. In the example, the com-

position and all activities require a strong traceabil-

ity policy, so all composition and service level events

are tracked and stored in the Tracking Data Reposi-

tory (a database). While the composition is running,

it is possible monitor its status in real-time. This is

achieved in the platform through the following steps

(see Figure 5):

Figure 5: Finding the current status of a running process.

(1) The Application calls the getStatus() operation

from the Traceability and Auditing Center

(TraceAudCenter) service;

(2) The TraceAudCenter informs the Real Time

Tracking Service (RealTTService) that a given

composition instance should be tracked following

a given policy;

(3) The RealTTService subscribes to the tracking

events generated by the WF Runtime Engine (Ex-

ecutionEngine) for the given instance. It uses a

custom proﬁle that implements the active trace-

ability policy (see also Section 3.6). From this

moment, the events speciﬁed in the proﬁle start to

be delivered to the RealTTService;

(4) The TraceAudCenter asks the RealTTService for

the status of a given composition (or also of an

activity or property);

(5) The status response (statusResp) is sent back to

the TraceAudCenter and then to the Application

(6).

The next time the application requests real-time sta-

tus information of the same composition instance, the

steps (2) and (3) do not need to be repeated (the ﬂow

starts then from step (4)) as the instance is already be-

ing monitored. It is also possible to audit the process

execution anytime after its conclusion using the Au-

diting Service, which offers to the applications a con-

venient way to access the tracking information stored

in the repository. Note, though, that only informa-

tion in accordance with the selected traceability pol-

icy will be present there.

3.6 Implementation Issues

The CoGPlat middleware infrastructure is imple-

mented with the support of the following technolo-

gies: the Service Bus exports traditional Web Ser-

vices (WSDL + SOAP/HTTP); the middleware core

runs over the Microsoft .NET framework and is writ-

ten in C#; the service discovery layer uses the

OWLS-MX matchmaker (Klusch et al., 2006); and

the composition executions are performed by the

Windows Workﬂow Foundation (WF) runtime engine

(Bukovics, 2007). The Traceability and Audit-

ing Center implementation relays on the WF Track-

ing Service functionalities. It provides three event

categories: Workﬂow Events which correspond to

changes in the composition instance status; Activ-

ity Events which correspond to changes in the ex-

ecution status of individual activities (services); and

User Events that can be generated at any point in the

life cycle of the composition instance (those events

are customized and deﬁned during the construction of

the abstract compositions). The default WF Track-

ing Service behavior generates events for all work-

ﬂow and activity types. Therefore, to implement a

behavior that corresponds to the chosen traceability

policy, CoGPlat deﬁnes three custom WF tracking

proﬁles, one for each of the possible traceability poli-

cies (strong-, weak- and zero- traceability). Note that

as the ﬁrst policy (strong) allows customization, ad-

ditional tracking behavior might be speciﬁed during

the construction of the abstract compositions. Besides

what is (and what is not) being monitored, another

important question is what to do with the information

ICEIS 2008 - International Conference on Enterprise Information Systems

188

produced by the tracking service. The Traceability

and Auditing Center uses the default WF SQLTrack-

ingService, which provides ready-to-use functionali-

ties to record data to a SQL Server database. This

recorded data can be later used for auditing purposes.

In addition, CoGPlat implements a custom tracking

service which allows the on-line monitoring of an ex-

ecuting composition.

4 CONCLUDING REMARKS

The proposal of effective solutions to increase the

public administration transparency is becoming a key

issue for the success of the new citizen-centric ser-

vices and applications. Considering the challenges in-

troduced by this new scenario, this paper contributes

proposing a strategy to monitor and audit composite

e-Government services. It includes the proposal of a

set of Traceability Policies, used to semantically an-

notate the services, specifying the desired traceabil-

ity behavior. The paper also contributes by present-

ing an application scenario and by discussing its im-

plementation, based on the Windows Workﬂow Foun-

dation Tracking Service. The work presented in the

paper is in the context of the Traceability and Audit-

ing Center, one of the core facilities of the Citizen-

oriented e-Government Platform (CoGPlat). Future

works may include the extension of the traceability

policies, the implementation of new applications and

a study on the effects of the tracking mechanisms on

the performance of time-critical compositions, not so

frequent in the e-Government domain. There are still

many challenges to be faced before a fully citizen-

centric e-Government becomes a reality. However,

the visioned changes create the expectations of a fu-

ture where public administration processes may be-

come really transparent and efﬁcient, stimulating the

continuous research efforts in this ﬁeld.

ACKNOWLEDGEMENTS

The authors would like to thank the Brazilian agen-

cies CAPES and CNPq for the ﬁnancial support. Ad-

ditional thanks to the Fraunhofer-FOKUS Institute

(Berlin, Germany).

REFERENCES

Barnickel, N., Fluegge, M., and Schmidt, K.-U. (2006). In-

teroperability in egovernment through cross-ontology

semantic web service composition. In Workshop Se-

mantic Web for eGovernment, Montenegro.

Berners-Lee, T., Hendler, J., and Lassila, O. (2001). The

semantic web. Scientiﬁc American, 284(5):34–43.

Bukovics, B. (2007). Pro WF: Windows Workﬂow in .NET

3.0. Apress.

Fileto, R., Liu, L., Pu, C., Assad, E. D., and Medeiros, C. B.

(2003). Poesia: An ontological workﬂow approach

for composing web services in agriculture. The VLDB

Journal, 12(4):352–367.

GOV3 (2006). Citizen Centric Government (White

Paper). The GOV3 Foundation - Intel,

www.intel.com/go/government.

IEEE (1990). IEEE Standard Computer Dictionary: A

Compilation of IEEE Standard Computer Glossaries.

New York, NY.

ISO (1994). ISO 8402:1994 - Quality management and

quality assurance - Vocabulary.

Jefferson, T. (1789). Personal communication to R. Price.

KBSt (2006). SAGA - Standards and Architectures for

eGovernment Applications - Version 3.0. German Fed-

eral Ministry of Interior, www.kbst.bund.de/saga.

Klusch, M., Fries, B., and Sycara, K. (2006). Automated

semantic web service discovery with owls-mx. In

Proc. of the 5th Intl. joint conference on Autonomous

agents and multiagent systems (AAMAS), pages 915–

922, USA. ACM.

Medjahed, B. (2004). Semantic Web Enabled Composition

of Web Services. PhD thesis, Virginia Polytechnic In-

stitute and State University.

Medjahed, B. and Bouguettaya, A. (2005). Customized de-

livery of e-government web services. IEEE Intelligent

Systems, 20(6):77–84.

Narayanan, S. and McIlraith, S. A. (2002). Simulation, ver-

iﬁcation and automated composition of web services.

In Proc. of the 11th Intl. WWW Conf., pages 77–88.

ACM Press.

Papazoglou, M. and Georgakopoulos, D. (2003). Service-

oriented computing. Communications of the ACM,

46(10):25–28.

Santos, I. J. G. and Madeira, E. R. M. (2006). Applying or-

chestration and choreography of web services on dy-

namic virtual marketplaces. International Journal of

Cooperative Information Systems, 15(1):57–85.

Santos, I. J. G., Madeira, E. R. M., and Tschammer,

V. (2005). Towards dynamic composition of e-

government services - a policy-based approach. In

5th IFIP Intl. Conf. on e-Commerce, e-Business and

e-Government (I3E), pages 173–185. Springer.

W3C (2005). Web Service Modeling Ontology (WSMO)

Primer. http://www.w3.org/Submission/WSMO-

primer/. Member Submission.

W3C (2007). Semantic Annotations for WSDL and XML

Schema (SA-WSDL). http://www.w3.org/TR/sawsdl/.

TRANSPARENCY IN CITIZEN-CENTRIC SERVICES - A Traceability-based Approach on the Semantic Web

189