Studies in Informatics and Contro l, Vol. 20, No. 2, June 2011 http:www.sic.ici.ro 129 [601662]

Studies in Informatics and Contro l, Vol. 20, No. 2, June 2011 http://www.sic.ici.ro 129
1. Introduction
The survival and development of Romanian
enterprises require alignment with
international norms on product and service quality. First and foremost, all top manufacturers and importers require that the products they are supplied with be produced
in systems that facilitate their improvement in
terms of quality, environment, health, and occupational safety, etc.
This paper aims to create/present a
methodology for the integration of the
enterprise and its departments as well as the
integration of the enterprise with its suppliers and customers in order to answer promptly to changes on the market. The model does not aim at developing a system able to be implemented – from scrap – into an
enterprise, but to focus on the integration of
the already extant systems (departments, human resources, software, hardware, etc.), because of the lower implementation costs than those involved in the case of a
specialized ERP systems, such as SAP (a
system which can be purchased by multinational enterprises only, because of its high costs).
2. Current Status
In Romania, the development of the infrastructure, of the communication and information systems and services, is a
mandatory requirement for the overall
economic development, as well as for the integration into globalized economy, and especially in the economic and social environment of the EU economy.

The approach is meant to support small and
medium-sized enterprises in their
development and in adapting to the new “e-business” and “e-commerce” environment, in achieving better communication standards within an organization, which is no longer confined within “four-walls”, but extends on
a national, European and worldwide level, in
using the web and Internet technologies which help (with low costs) Romania meet the recent EU requirements related to an “e-Europe” [1], [13].
3. The Study of the Modeling
Languages through Multi–
attribute Decision Analysis
In order to perform a multi-criterion analysis
we have used the methodology approached in the comparative analysis of reference architectures presented in [3], [5], [8].
For this multi-attribute analysis we selected
the following modeling languages:
GRAI/GIM, Petri Nets, OOA/OMT, CIMOSA, IDEF0, IDEF1X, IDEF3, IEM, ARIS, EXPRESS and UML.
These representative techniques and methods
are presented below.
GRAI and GRAI/GIM (Graph with Results
and Activities Interrelated/GRAI Integrated Methodology). Its proposed methods are static descriptive models [7];
Petri Nets . These [11] are graphical and
mathematical tools for representing and
analyzing discrete event systems, characterized by parallelism and synchronization. Modeling enterprises by Web Technologies and Multi-criteria Analysis Used in
Enterprise Integration
Marius CIOCA1, Lucian – Ionel CIOCA1, Luminita DUTA2
1 Lucian Blaga University of Sibiu,
10, Bd-ul Victoriei, Sibiu, 550024, ROMANIA
2 VALAHIA University of Targoviste,
2, Bd-ul Carol I, 130024, Targoviste, ROMANIA
Abstract: The studies, research and solutions presented below started from the acute need of small and medium-size d
enterprises to integrate their processes, services and rela tionships with customers and suppliers, in a cost-effective
manner, given the fact that ERP systems ar e quite expensive for such organizations.
Keywords: Web technologies, multi-criterion analysis, m odelling languages, enterprise integration.

http://www.sic.ici.ro Studies in Informatics and Contro l, Vol. 20, No. 2, June 2011 130using Petri networks can be done from the
following perspectives: functional, informational and resource using extensions of these networks. A very good example in this regard is the work of Neagu [9] which presents formalism and the use of the Petri
networks in detail;
OOA (Object Oriented Modeling) and
OMT . Object-oriented methods are
increasingly used in enterprise modeling [5];
CIMOSA (Computer Integrated
Manufacturing – Open System Architecture)
covers, in addition to f unctional, informational
and resource aspects the organizational aspects of the enterprise at different levels, thanks to the detailed model [14];
IDEF0 (Integrated Definition). The IDEF0
Language is based on the SADT (Structured
Analysis and Design Technique) technique, developed by Douglas T. Ross and SofTech, Inc. and it is used for modeling functional aspects. Its aim was to develop generic
subsystems to be used by enterprises in order
to achieve a substantial improvement of the the industry as a whole [4];
IDEF1X . The Informational Modeling
technique IDEF1X is used to model data in a standard, consistent, predictable manner, to
manage data as a resource;
IDEF3 is a graphical modeling method used
to facilitate the description of processes. It is composed of a limited set of constructions to describe the steps of the process (termed as
units of behavior) and their connections. The
resulted models are static descriptive models. However, they can be expanded and translated into Petri networks for analysis and simulation purposes;
IEM (Integrated Enterprise Modeling) covers
the informational and functional perspectives of the enterprise. It is based on an object-oriented modeling approach, in which the central structure is a IDEF0 block in which all inputs and outputs are state objects of three
types: commands, products and resources;
ARIS (Architecture for Information System)
is a product of Prof. Scheer GmbH, modeling the enterprise from a functional, informational and organizational perspective; EXPRESS is based on Chen's ER model
(Entity Relationship) to specify STEP entities (ISO 10303-11:1994);
UML (Unified Modeling Languages). The
unified modeling language is a language for building, specifying, visualizing and
documenting software products. It is
supported by numerous CASE tools, of which Rational Rose is the most representative. Because it is a simple and expressive language, of visual modeling, independent of
programming languages, flexible, and
because of the growing market for OO (object oriented) tools, this language is becoming increasingly used in modeling systems as well as enterprise modeling [2].
Thus, in order to select the best modeling
language recommended for the EMLs
(Enterprise Modeling Languages) block of the GERAM architecture, we have analyzed, based on the literature [4 ], [11], [12], [14]
etc, the following modeling languages
GRAI/GIM, Petri networks, OOA/OMT,
CIMOSA, IDEF0, IDEF1X, IDEF3, IEM, ARIS, EXPRESS and UML which form the alternative set A={A1, A2, , A11 }. [6], [7].
In order to establish the evaluation criteria
presented below, we considered the modeling
requirements that modeling languages have to
meet. [9], [10]. Thus, for the selection of the modeling language based on the multi-attribute analysis 10 criteria have been defined, i.e. the set E={E1, E2, …, E10} as follows:
 From the point of view of the modeling
and simulation support:
◦ E1 – CASE tools;
◦ E2 – methodological support;
 From the point of view of the
modeling perspectives:
◦ E3 – the function perspective;
◦ E4 – the information perspective;
◦ E5 – the resource perspective;
◦ E6 – the organization perspective;
 From the point of view of the modeling
and coverage principles:
◦ E7 – distinguishing between problems;
◦ E8 – functional decomposition;
◦ E9 – genericity;
◦
E10 – distinguish between functionalities;

Studies in Informatics and Contro l, Vol. 20, No. 2, June 2011 http://www.sic.ici.ro 131The weight for every criterion is in this order
{0.20, 0.20, 0.10, 0.10, 0.10, 0.10, 0.5, 0.5, 0.5, 0.5}, where the set of weights form vector w, which meets the requirement: the sum of all weights is 1.
The scores employed in the multi-attribute
analysis are cardinal, and range between 1
and 5.
The table of consequences resulted after the
use of the established score for each criterion, in the case of each alternative is shown in the
Table 1.

The decision table is obtained by normalizing
the scores presented above which is to be made through formulas 2 and 3, and the final score obtained by each alternative is calculated according to formula 1.

n
jjj i rw J
1, (1)
where: w is the weight vector w j of criteria E j
(j=1,1,nc), nc is the number of evaluation
criteria and r j represents the normalized
scores according to the formulas below:
jij Mj
ijss sr) (
, (2) jmj ij
ijsssr) (
, (3)
where: sj is the variation range defined by
extreme scores (s mj and s Mj) of the
alternatives considered, and is calculated
according to the formula:
mj Mj j s s s , (4)
Pairs (s mj and s Mj) of the lower values and the
highest considering criterion E j are:

s
mj = min(s ij,i = 1,2,…,na); j = 1,2,…,nc (5)
sMj = max(s ij,i = 1,2,…,na); j = 1,2,…,nc (6)
where: na is the number of alternatives, and
nc is the number of evaluation criteria.
The graphic representation of the results in
the table above is shown in Figure 1.
As shown above, the UML was ranked
second. However, we have selected and recommended this language because of the following reasons:
 the UML integrates more easily with
Web technologies used for the development of the proposed model. Thus, the data may be modeled by means

http://www.sic.ici.ro Studies in Informatics and Contro l, Vol. 20, No. 2, June 2011 132of the UML and handled by means of the
XML. In addition, it enables the user the generate Java code (program code);
 IT specialists dealin g with industrial
systems analysis and development are more “inclined” to work with the UML, unlike system analysts coming from other fields
and who probably prefer IDEF or ARIS.

4. The Study of Information
Systems Reference Architectures
in the Industrial Environment
The study of the enterprise integration
reference architectures through multi-
attribute decision analysis may be employed
by means of the method presented above and described in [3]. From all architectures in the
literature (GERAM, CIMOSA, GRAI/GIM, PERA etc) it is recommended that GERAM
be used for enterprise modeling.
CIMOSA In 1985, the ESPRIT program
through the AMICE consortium lays the foundation for a CIM architecture for enterprise integration – CIMOSA (Computer
Integrated Manufacturing – Open System
Architecture). CIMOSA has three components:
 the enterprise modeling framework;
 the modeling language;
 the enterprise infrastructure.
CIMOSA does not only provide a reference
model, but follows the modular model
constructed in the integrated enterprise
engineering environment . In a step by step
approach, the degree of abstraction of the model is reduced from the general (generic) level, which is specific to any enterprise at an intermediate level, specific to
a particular sector and, ultimately, to the lowest (specific) level at which the model refers to a particular enterprise.
PERA Purdue Enterprise Reference
Architecture (PERA) was designed to assist the
industry in its development and implementation
efforts of integrated production systems. It is a

complete and widely documented
methodology to define, design, build, install and operate an integrated enterprise system or a major automation project and was developed by Purdue University, USA, in
collaboration with a consortium of industrial
companies. This architecture is known in particular because of its main focus on the human aspect and its interactions with various aspects of automation.
GRAI/GIM (Graph with Results and
Activities Interrelated/GRAI Integrated
Methodology) was developed by the laboratory of the Bordeaux University (France). The reference model GRAI considers the production system as a set of
three subsystems:
 the decision-making system;
 the information system;
 the physical system.
GRAI / GIM covers the entire life cycle of
the production system except operation and decommissioning phases. The four perspectives differ from those of CIMOSA
architecture, a decision-making and a
physical perspective are introduced. The four perspectives used duri ng the user-oriented
analysis and design are converted into three perspectives of implementation during the
technical-oriented design stage.

Figure 1. The classification of the modeling languages resulted from the decision table

Studies in Informatics and Contro l, Vol. 20, No. 2, June 2011 http://www.sic.ici.ro 133IEM (Integrated Enterprise Modeling) [4] is
a reference architecture that covers the functional and informational perspectives of the enterprise. It is based on an object-oriented modeling method in which the central structure is an IDEF0 (Integration
DEFinition) block in which all inputs and
outputs are object states of three categories:
 – commands;
 – products;
 – services.
ARIS performs a detailed description of an
enterprise in terms of data model, functional, organizational and process models, as well as
from the perspective of model analysis, it is
an indispensable requirement for defining terms discussed in an enterprise. Some of these terms are:
 business process reengineering;
 effective selection and introduction of
standard application systems;
 development of specialized software;
 development of architectural plans for
information technology;
 documentation on the existing data
processing method;
 workflow management.
ARIS architecture (Architektur integrierter
Informationssystem) is strongly supported by software tools such as ARIS ToolSet, which
includes components of modeling, analysis
and navigation of user interfaces, functions and integrated database. Some of these features are:
 multi-user operating capacity;
 graphical user interface, object-oriented,
under a Windows environment;
 relational database with automatic restoration;
 SQL/ODBC interface;
 management of reference models;
 assessments and comparisons
between models;
 adapting the model, based on rules
(starting from the reference model);
 easy display of the entire available
information for each group of objects and for individual objects;  wide range of browsing functions for
complex enterprise models.
GERAM . Starting from the assessment of the
most representative reference architectures for enterprise integration (CIMOSA, GRAI/GIM and PERA), the IFAC / IFIP workgroup have
developed a global defini tion of a generalized
architecture. The proposed framework was
named GERAM (Generalized Enterprise Reference Architecture and Methodology). GERAM refers to those methods, models and tools that are necessary to build and operate
the enterprise.

Figure 2. Elements of enterprise integration
GERAM provides a description of all the
elements recommended in engineering and enterprise integration, thus establishing a
standard for the collection of tools and methods enterprise will benefit from in order
to successfully manage initial integration
design and of the change processes that may occur during life cycle of enterprise operation.
GERAM does not impose a particular set of tools and methods, but defines the criteria that
must be met by such a combination.
5. Conclusions
In conclusion, we have come to the following:
Where, in order to cut costs, a small or
medium-sized enterprise should use [2]:
 the GERAM architecture for enterprise
modeling;
 the UML is the recommended language for
enterprise modeling and for “filling” the

http://www.sic.ici.ro Studies in Informatics and Contro l, Vol. 20, No. 2, June 2011 134EMLs (Enterprise Modeling Languages)
block in the GERAM architecture;
 the recommended ISO standards should
be taken into account during the development of the enterprise integration model/methodology;
 For the integration of the enterprise both
within its “four walls”, but especially
with its suppliers and customers as well
as with its subsidiaries, opensource Web technologies should be used, due to the fact that they involve not costs (licenses are free) and they are flexible and
established. From this point of view, we
recommend the following web technologies: the PHP application server, the MqSQL/PostgreSQL database server, Java, JavaScript, the Apache Web server
running on Linux as well as the XML
language for data handling within and outside the organization.
REFERENCES
1. BURAGA, S. C, M. CIOCA, Using
Semantic Web Technologies to
Enhance the Inter-Connectivity Between the Components of an E-
Learning System , 4th Intl. Conf. on
RoEduNet Romania, May 20-22, 2005
Targu Mures ROMANIA, pp. 17-18, ISBN: 978-973-7794-29-1;
2. CIOCA, M., S. C. BURAGA, New Tools
for Human Resource Management in
e-Business: Combining UML
Language, Reference Architectures
and Web Programming , IEEE Intl.
Conf. on Industrial Informatics, Aug. 21-24, 2003 Alberta CANADA, pp. 170-
174, ISBN: 0-7803-8200-5;
3. CIOCA, M., LI. CIOCA, Multi-criterion
Analysis of Reference Architectures
and Modeling Languages used in
Production Systems Modeling , 3rd
IEEE Intl. Conf. on Industrial Informatics
(INDIN), Aug. 10-12, 2005, Perth
AUSTRALIA, pp. 230-233, ISBN: 0-7803-9094-6;
4. FILIP, F. G., B. BARBAT, Informatica
industriala: Noi paradigme si aplicatii ,
Editura Tehnica, Bucuresti, 1999; 5. FILIP, F. G, G. NEAGU, Object
Oriented Approach to Software
Engineering for CIME , in P. Groumpos
(Ed.). Preprint, 8th IFAC/IFIP/IFORS/IMACS/IFIP Symp. “Large Scale Systems: Theory and
Applications”, Patras (July 15-17), 9,
1994, pp. 400-404;
6. ISO 10303-11:1994 Industrial
Automation Systems and Integration –
Product data representation and
exchange. Part 1: Overview and
fundamental principles;
7. KOSANKE, K., Comparison of
Enterprise Modeling Methodologies . in
J. Goosenaerts, F. Kimura, H. Wortmann (Eds.) Information Infrastructure Systems
for Manufacturing. Champman & Hall,
London, 1997, pp. 115-127;
8. KROGSTIE, J., S. ARNESEN, Comparing
Languages for Enterprise Modeling using
a Language Quality Framework ,
Norwegian University of Science and
Technology, Institute of Computer and Information Sciences, 2003;
9. NEAGU, G. Contributii privind
conceptia si metodele de rezolvare a
sistemelor pentru conducerea
atelierelor de fabricatie , Teza de
doctorat, Universitatea Politehnica din Bucuresti, 1998;
10. OPREAN, C., C. V. KIFOR, Modelling
the Implementation, Usage and
Maintenance of the ISO 9000 Quality
Management System using the
Reference Architectures . in TMCM
1998, Iasi, Romania;
11. REISIG, W. Petri Nets . Springer Verlag,
Berlin, 1985;
12.
RUTAKYAMIRWA, N. N., Modeling of
Business Process in Distributed
Environment : a Case Study of Issuance of
Certificate of Occupancy in Tanzania, 2002;
13. SUDUC, A., M. BIZOI, F. FILIP, User
Awareness about Information Systems
Usability , Studies in Informatics and
Control, Vol. 19(2), June 2010;
14. WILLIAMS, TJ et al, Architectures for
Integration Activities and Enterprises ,
Computer in Industry, 24, 1994, pp. 111-139.