Workflow modelling within selected departments of the public administration focused on the regional office

Workflow modelling within selected departments of the public administration focused on the regional office

Jan Capek; Robert Bata

University of Pardubice, Pardubice, The Czech Republic

^{Address for correspondence}

ABSTRACT

E-government means more likely transformation in nature, affecting the management of human, technological and organisational resources processes. Some authorities state that an integrated process model should make it easy to identify possible ways for improvement. Instruments like Petri-Nets can be suitable for partial solutions in public administration. The main advantages of Petri nets are their graphical notation, their simple semantics, and the rich theory for analyzing.

Keywords: Workflow modelling, Petri nets, public administration, e-government.

1 INTRODUCTION

E-government is becoming a global phenomenon that is an influential topic for politicians, policy makers as well as ordinary citizens. The World Public Sector Report of the United Nations indicated that in 2003, more than 173 countries had prepared government web sites. E-government is predicated on leveraging the capabilities and power of ICT to deliver services provided by governments at local, municipal, state and national levels. Beyond service delivery, e-government offers additional channels of interaction among governments, businesses, and citizens, separately or collectively (UN, 2003). However, e-government is more than a technical solution. It more likely means transformation in nature, affecting the management of human, technological, and organisational resources processes. In literature there is a number of different definitions of e-government. Some of them are rather narrow, focusing on using ICT, particularly the Internet, as e.g. "the use of technology to enhance the access to and delivery of government services to citizens, business partners and employees", (Deloitte, 2004). Others view e-government more broadly like transformation of government as such. Examples can be:

Electronic information-based services for citizens (e-service) with reinforcement of participatory elements (e-democracy) to achieve objectives of balanced e-government (Bertelsmann Foundation , 2001)

The use of information and communication technologies, particularly the Internet, as a tool to achieve better government (OECD, 2003).

The use of ICT in public administration combined with organisation changes and new skills, in order to improve public services and democratic processes and strengthen support to public policies. COM (2003).

The use by the government of Web-based Internet applications and other ICTs, combined with processes that implement these technologies, to

1) enhance the access to and delivery of government information and services to the public, other agencies, and to government entities; or

2) bring about improvements in government to operations that may include effectiveness, efficiencies, service quality, or transformation" (US government, 2002).

These definitions may be useful in describing e-government in a broad-based manner, but don't help follow some deeper issues and considerations relating to the construct, and fail to capture the more complex aspects of transforming government or acknowledging the role of the ICT elements.

Consequently, most implementation activity centres on service delivery dedicate little emphasis on real transformation of the services themselves or the processes associated with their delivery (Grant, Gerald and Chau, Derek, 2005).

The next claim 'any conceptualization of e-government' needs to address a variety of concerns beyond the service delivery elements. Based on a comprehensive literature review, they suggest this definition. In the work on eGovernment in the EU Commission, they focus on these overall objectives (COM, 2003):

• A public sector that is open and transparent, that is understandable and accountable to the citizens, and open to democratic involvement and scrutiny.

• A public sector that is at the service of all, being inclusive and exclude no one from its services.

• A productive public sector that delivers maximum value for taxpayer's money.

Departing from these clearly defined goals and priorities a simpler framework is proposed, defining basically three variant groups of stakeholders: politicians, public institutions, citizens, businesses and civil society and thereby distinguishing between 3 different dimensions (Grønlund, 2002):

1) the democratic dimension, focusing on political processes and interaction between the constituents and the government,

2) the service dimension which comprises the delivery of all types of electronic services, and

3) the administrative dimension including various types of management work, internal routines etc. This may be illustrated in the following way (Thomas et al., 2004; Jansen, 2005) and we are focused on the e-service part:

2 GOVERNMENT PROCESSES MODELLING

Outstanding results of customer service orientation are the installation of call-centres, service pages, and the connection with quality management. But, the change in internal processes will not be easy since the state cannot exclusively focus on the customers' needs and benefits. The internal workflow of public authorities is governed by a legal framework. Furthermore, there are hundreds of decrees for almost each administrative process. These decrees seldom contain any instruction about the output, but they usually regulate how to create a required one. The following aspects have to be taken into account as a result when modelling government processes (Wimmer, Traunmüller, 2003):

Organizational and technical perspectives are not easy to synchronize within a project since small changes on the one perspective may require big efforts on the other.

Internal and external views on the workflows differ considerably. Public online services offered in portal solutions should use a customer oriented workflow; This is a perspective in which the state has little experience. Thus, current laws do rarely support efficient and effective workflows.

E-Government modelling methods are called to respect the characteristics of public services. While most of the existing modelling methods were developed in order to optimize supply chains or production, the goals in the public administration are usually defined in a different way.

The synchronisation of political and strategic perspectives is imported. Political influence and its dependency on election terms have to be understood and need to be well separated from the long term strategic perspective of modernizing the state's processes.

Each of these aspects has to be analyzed individually for every country at a more detailed level. Laws and decrees are often the major barrier to re-engineering public workflows (Lenk, Traunmüller, Wimmer, 2002).

Each step of an administrative process has to be analyzed and, if necessary, modified (Lenk, Traunmüller, 1999). Another reason for neglecting process modelling is the fact that the main goal has been to offer information and communication online. This could be a good introduction but will not be enough to fulfill the high expectations in e-Government. To provide services at the transaction level, it is crucial to remodel public processes towards e-Government. A lot of methods exist for the modelling of business processes. Some of these methods are used to represent the processes embedded in major enterprise resource planning software (e.g. SAP R/3, ARIS...). In principle, these methods can be used to model government processes as well (Becker, 2004).

Some authorities have declared that an integrated process model should make it easy to identify possible ways for improvement. It should be easy to understand and hence to be accepted by the authorities. Highly formal modelling languages like Petri Nets or the Unified Modelling Language do not seem suitable for e-Government (Thomas et al., 2004). The modelling approach should be independent of the primary purposes (e.g. customers needs, higher efficiency, cost reduction, etc.) of the re-engineering efforts (Šimonová, Lešák, Kalhous, Vávra, 2007), (Šimonová, Naiman, Kopackova, Bilkova, Jonasova, 2007). But other authors have different opinions and experiences.

The Semantic Process Language (SPL) is a formal language suitable for the specification of process sets and their verification concerning process properties (Simon, 2006). It is a novel language based on Petri nets (Petri, 1962), (Reisig, 1986) - especially a variation of Petri nets called Workflow nets (Aalst, 2007) - and previous work on a Logic of Actions (Genrich, 1978; Lautenbach, Simon, 2000; Simon, 2002). Opposite to other formal languages, the meaning of the SPL words (which are called modules) is not defined by grammar but by building rules which take modules as input and synthesize a special class of Petri nets (called Module nets). Specific firing sequences of these nets are then interpreted as the (sequential) processes defined by a module.

The Semantic Process Language is enriched by the specification of the resources activities use and by the possibility to connect modules in order to find similarities between process sets. This last property makes SPL particularly applicable to the problems of improving the quality of public processes which are discussed in this paper.

Petri nets are a well-accepted formalism for modelling concurrent and distributed systems in various applications like Workflow management, embedded systems, production systems, or traffic control. The main advantages of Petri nets are their graphical notation, their simple semantics, and the rich theory for analyzing.

Petri net consist of the places p denoted as Ë%, transitions t denoted as ^____ and directed arcs (connections) denoted as →.

3 OUR APPROACH

We have focused our research on E-services (see fig. 1). The elaborate information flow model is described in Figure 3. Individual symbols represent the pages of applications that go through the modelled department. Initial place is the transition t_1, generating the processes (in this case it is the pages of incoming documents) into the system. Place p₁ serves as a "waiting room" of the documents to be further processed. Transition places t₂ - t₇ withdraw the delivered documents. The transition places represent the moment of a document assignment to the appropriate department. Transition performance is determined by the probability of 16.6%. This value is determined by the number of departments of the appropriate division and will be different at the divisions with different number of subdepartments. The model describes the Department of environment and agriculture of the Regional Office of Pardubice with 6 subdepartments, therefore the value reaches 100/6=16.666.

An essential prerequisite is equal load of all the departments, which corresponds to the real state. Assignation of symbols to places p₂ - p₇ presents the moment of a document processing. Transitions t₁₄ - t₁₉ and analogically t₂₀ - t₂₅ up to t₄₄ - t₄₉ demonstrate the moment when a department has finished processing an application and transferred it onto another one. Performance of the transitions is determined by the probabilities. The probabilities of the individual transitions performance are determined by a table of probabilities indicating the number of documents and the number of the involved departments. To provide for a document processing by an accurate number of departments, different colours are used to symbolize transitions t₁₅ - t₁₉, analogically at t₂₁ - t₂₅...., t₄₅ - t₄₉. The colours correspond to the colour of an arc that comes out of the appropriate transition. Therefore a symbol, with the probability of 15% (an application processed by one department) continues to p₉, and with the probability of 85% to p₈, with an assigned colour. If a symbol gets through a transition t₁₉, t₂₅, t₃₁, t₃₇ or t₄₉, the colour orange is assigned and continues via p₈ into one of the transitions t₈ - t₁₃. Transitions t₈ - t₁₃ are of predicate type; the predicates prevent an application from going into one department more than just once. Symbols go through one of the transitions p₂, p₃,...., p₇, and via an orange arc they enter an appropriate transition and from that places to the transitions t₅₄, t₅₉,..... t₇₉ where the arc colour changes into yellow. The symbol continues to p₈, and further to t₈ - t₁₃ the same way. Analogically, it goes through one of the places p₂ - p₇. The colour gradually changes from orange to yellow, blue, green and red. The red colour symbol then continues, upon making an appropriate transition, into place p₉ as a black colour symbol. Transitions t₁₄ - t₇₉ are of a time type, their period of life is calculated from empiric data. A model, graphically expressed by Petri nets tools, is described in Fig. 2.

As it is seen in Fig. 2, such concept is rather complicated. The office permeability in terms of an application processing and its mutual cooperation is the monitored events, therefore the model in Fig. 2 involved redundant information concerning a particular link of a processed application to an appropriate department. As far as time measuring is concerned, the question of a department to process an application is irrelevant; it is the processing time that matters. On condition that we leave out the requirement to describe the link of a processed document with the organizational structure of the office, the whole procedure would be significantly simplified, not limiting its information value, because we observe the speed of processing within the division. The information about the type of department is redundant. To be implemented in selected environment, a simplified model can be used, that does not contain such information. Then, the first department described in the model does not correspond to a particular department, but to the first department to process an application. It can be the department of integrated prevention, for example, or the water management department, as the case may be. Due to this simplification we do not need the colour marking in Petri nets, nor do we need the use of the predicates. Time Petri net is sufficient, involving the defined probabilities. A simplified model is shown in Fig. 3.

Table 3

Transitions t₂ - t₇, are indicated in this model, or their equivalents, marking the individual departments of the division of environment and agriculture of the Regional Office of Pardubice. Place p₁ represents the department which was involved first; p₂ is the second in the given order, etc.

It is essential to integrate the time aspect, so that the model really describes the course of an application being processed by a monitored department. The performance of transitions t₂, t₄, t₆, t₈, t₁₀ and t₁₂ must be determined by the time of processing. The appropriate values are then adjusted to the requirements of a given environment in which the model was applied. The given model is sufficiently simplified to be easily transferred into software tools and further processed.

The simplified model indicated in Fig. 3, implemented by means of the HPSim programme and graphically expressed in its environment is in Fig. 4

The simulation results could be exported from HPSim into MS Excel via an output file for further processing. The simulation starts with the initial marking of 15 tokens in p₂ and 0 in other places. A first step of simulation represents abnormal conditions until a queue of tokens in p₇ - p₁₁ is reached for each department. Every simulation step is in progress in time. To define reliability of simulation results it is necessary to determinate how many steps of simulation must be effected. The required number of simulation steps and the 95% reliability interval per 128 time units were determined experimentally for this model. The accuracy of the results and the reliability interval are indicated in Fig. 5. Simulations, however, were carried out for a higher number of steps, due to a better comparability of the model outcomes with the empirical data.

To verify the model accuracy, check up if data has been used. Graphic comparison of the empiric data and modelled values are displayed in Fig. 6.

The graph lists the trend-connections equations. The first equation at the top belongs to the empirical data, the second to the model outcome. Line 1 (purple points) describes the model outcome, line 2 (blue points) describes the empirical data. As the trend connections show, the model successfully reflects the real state. Both lines are almost parallel, the modelled data deviation is very small; it is propably caused by a small number of check up data. The comparison of the empirical data with the modelled values demonstrates that the model reflects the real state with sufficient accuracy. The figure clearly shows that the model respects the distribution of empirically collected data, which grows with the increasing number of pages in an application. Such distribution increase complies with the assumptions about the application processing procedure.

The transparency calculated from a simplified model of the existing state for 492 simulation steps at 0.568 pages/day can be, in relation with the 95% reliability interval, interpreted as a value deviating from the real value by ± 0,000284 pages/day. The 492 simulation steps have matched in this event for 3 years. Such an exactness for use in public administration is more than necessary, because the interpretation of this value can be, that the model can calculate with exactness ± 0,5 letter for one page as well. But this exactness will be reached with only 128 simulation steps, what represents 9,3 months. So that can be the model used for calculations in relatively short time periods with very good results.

4 CONCLUSIONS AND FUTURE WORKS

The results of our experiments show that the model successfully reflects the real state. Both lines (purple and blue) are almost parallel.indicating that the modelled data deviation is very small. It is probably caused by a small number of check up data.

The comparison of the empirical data with modelled values demonstrates that the model reflects the real state with sufficient accuracy.

The figure clearly shows that the model respects the distribution of empirically collected data, which grows with the increasing number of pages in an application. Such distribution increase complies with the assumptions about the application processing procedure.

The simulation models converge in a relatively fast manner to the stable points and, from the exactness point of view for the application within the public administration offices, it is more than sufficient.

This model will have to be improved, in the future, taking in the account time outgoings for necessary internal processes.

5 BIBLIOGRAPHY

Robert Bata

Currently working as a Assistant Professor at the Faculty of Economics and Administration

University of Pardubice, Pardubice, The Czech Republic

He received his Ph.D. degree at the Faculty of Economics and Administration

University of Pardubice, Pardubice

E-mail: [e-mail]

Institut of System Engineering and Informatics

Faculty of Economic and Administration

University of Pardubice

Studentska 95, 530 10 Pardubice

Manuscript first received: 30/1/2007

Manuscript accepted: 31/06/2007

• AALST, W. M. P., van der. The Application of Petri Nets to Workflow Managemen [on line]. 1998 [05.01.2007]. < http://is.tm.tue.nl/staff/wvdaalst/publications/p53.pdf >
• BECKER, J. et all. Prozessorientierte Reorganisation in öffentlichen Verwaltungen - Erfahrungen bei der Anwendung eines Referenzmodells, MobIS [on line]. 2004 [04.06.2004] <http://kom.wi-inf.uni-essen.de/mkwi04/teilkonferenzen_detail.htm>
• BERTELSMANN FOUNDATION. Balanced e-government - Connecting efficient administration and responsive democracy [on line]. 2001 [5.2.2001] http://en.bertelsmann-stiftung.de/index.html .
• COM Communication from the Commission to the Council [on line]. 2003 [5.12.2003] http://europa.eu.int/information_society/eeurope/2005/doc/all_about/egov_ communication_en.pdf
• DELOITTE RESEARCH PUBLIC SECTOR INSTITUTE. At the dawn of e-government. The citizens as customer [on line]. 2004 [04.06.2004] http://www.deloitte.com/dtt/article/0,1002,sid%253D37085%2526cid%253D60672,00.html
• GRANT, GERALD AND CHAU, DEREK. Developing a Generic Framework for E-Government. Journal of Global Information Management [on line]. , 13(1), 1-30, Jan-March 2005.http://www.igi-online.com/downloads/pdf /ITJ2725_dGLKgMR9SK.pdf .
• GENRICH, H.. Ein Kalkül des Planens und Handelns. In: Ansätze zur Organisationstheorie rechnergestützter Informationssysteme Oldenbourg Verlag, 1978. GMD Bericht 111, pp. 77-92.
• GRØNLUND, Å. Eds. (2002) Electronic Government: Design Visions and Management 1st printing, 2002. Idea Group Publishing, 2002. 388 pages. ISBN: 1-930708-19-x.
• LAUTENBACH, K., SIMON, C. Verification in a Logic of Actions. In: Fachberichte Informatik, No. 7-2000: 7. Workshop Algorithmen und Werkzeuge für Petrinetze. Koblenz: Universität Koblenz-Landau. 2000. p 19-24.
• LENK, K.; TRAUNMÜLLER, R.; WIMMER, M.: The Significance of Law and Knowledge for Electronic Government. In: Grönlund (Ed.) Electronic Government - Design, Applications and Management New York: Idea Group Publishing. 2002. p. 61-77.
• LENK, K.; TRAUNMÜLLER, R.: Öffentliche Verwaltung und Informationstechnik - Perspektiven einer radikalen Neugestaltung der öffentlichen Verwaltung mit Informationstechnik, Verwaltungsinformatik Nr. 20 1. printing. Heidelberg: Marcel Decker Verlag, 1999.
• OECD. The case of e-government: Experts from the OECD Report "The E-government imperative." OECD Journal on Budgeting. 2003, 3(1) p. 62-96. ISSN 1608-7143.
• Thomas O., et. al. EPK-Referenzmodelle für Verwaltungsverfahren. In: M. Nüttgens, F. Rump (Hrsg.) EPK 2004 - Geschäftsprozessmanagement mit Ereignisgesteuerten Prozessketten, 3. Workshop der Gesellschaft für Informatik e.V. Luxemburg: Institut für Wirtschaftsinformatik, 2004. p.39-54.
• PETRI, C. A. Kommunikation mit Automaten, Schriften des Institutes für instrumentelle Mathematik. Bonn. 1962.
• REISIG, W. Petrinetze: eine Einführung, 2. printing. Berlin: Springer-Verlag 1986. ISBN 3-540-16622-X.
• SIMON, C. A Logic of Actions to Specify and Verify Process Requirements. In: The Seventh Australian Workshop on Requirements Engineering. Melbourne, Australia: AWRE 2002.
• SIMON, C. Integration of Planning and Production Processes, Mathmod 2006, Special Session: "Petrinets: Current Research Topics and their Application in Traffic Safety and Automation Engineering", Wien, Österreich, (accepted).
• UN. UN Global E-government Survey [on line]. c2003 [08.12.2003]. http://unpan1.un.org/intradoc/groups/public/documents/UN/UNPAN016066.pdf .
• US GOVERNMENT. The e-government act of 2002. HR 2458 [on line]. 2002. http://csrc.nist.gov/policies/HR2458-final.pdf .
• WIMMER, M.; TRAUNMÜLLER, R. Geschäftsprozessmodellierung in e-Government: eine Zwischenbilanz, eGov days 2003 Arbeitskreis Organisation [on line]. 2003. http://falcon.ifs.uni-linz.ac.at/eGovProzessmodellierung/wimmer_traunmueller.pdf (20-10-2004).
• IMONOVÁ, S., LEÁK, J., KALHOUS, M., VÁVRA, M. Current approaches to data modelling. In Scientific Papers of the University of Pardubice Series D Pardubice: Univerzita Pardubice, 2006. Series D, Fakulty of Economics and Administration, 11 (2007), ISBN 978-80-7194-936-7, ISSN 1211-555X, s. 126-132.
• IMONOVÁ, S., NAIMAN, K., KOPACKOVA, H., BILKOVA, R., JONASOVA, H. Easy Business Intellingence Tools for Data Analysis. In Transactions on Communications Venice: WSEAS-Press, 2007. Issue 1, Volume 6, January 2007. ISSN 1109-2742, p. 250-255.
• JANSEN, A. E-government- why and what. In. Tessem,B.J., Iden og Cristiensen, G. (ed.) NOKOBIT 2005 21-23.11.05 Bergen , ISBN 82-8033-026-7

Publication Dates

• Publication in this collection
  11 Mar 2011
• Date of issue
  2007

History

• Received
  30 Jan 2007
• Accepted
  31 June 2007