<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Integration of Business Modeling and IT Modeling</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Girts KARNITIS</string-name>
          <email>Girts.Karnitis@lu.lv</email>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Janis BICEVSKIS</string-name>
          <email>Janis.Bicevskis@lu.lv</email>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Jana CERINA-BERZINA</string-name>
          <email>jana.cerina_berzina@di.lv</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Datorikas Instituts DIVI</institution>
          ,
          <addr-line>A. Kalnina str. 2-7, LV-1050 Riga</addr-line>
          ,
          <country country="LV">Latvia</country>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>University of Latvia</institution>
          ,
          <addr-line>Raina 19, LV-1586 Riga</addr-line>
          ,
          <country country="LV">Latvia</country>
        </aff>
      </contrib-group>
      <fpage>81</fpage>
      <lpage>93</lpage>
      <abstract>
        <p>The paper covers our research in improving business and IT process modeling. Practical experience shows that universal modeling languages such as UML and BPMN are IT oriented and are difficult to accept by business people. Business models need to use such hard to formalize business terms as legal acts, informal activities descriptions and comments etc. We propose novel approach how to create or choose modeling tools according to model usage. We propose 4 usages for modeling: model to understand business processes, model for business process definition, model as information system specification and model that can be interpreted by computer. All these four models can be perceived as four highly interconnected parts of one single model. We propose to create unique domain specific language for each model. We analyze functions specific for each model type. Authors last six years practical experience for creating and practical introducing related business and workflow oriented information system models are analyzed.</p>
      </abstract>
      <kwd-group>
        <kwd />
        <kwd>Business process modeling</kwd>
        <kwd>domain specific languages</kwd>
        <kwd>modeling</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>Introduction</title>
      <p>
        Business process modeling and information system specification have number of
problems irrespective of wide research in a field of Model Driven Software
Development (MDSD) [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ] and Model Driven Architecture (MDA) [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ]:
•
      </p>
      <p>
        Universal modeling languages UML and BPMN [
        <xref ref-type="bibr" rid="ref3 ref4">3, 4</xref>
        ] do not provide
language easily understandable by both – IT specialists and business people.
IT specialists propose to use UML than can describe operation of IS objects.
In such case process part outside the computer stays weakly specified.
      </p>
      <p>Therefore business people cannot adequately describe their requirements.
•</p>
      <p>Model syntax prevails over model’s meaning – semantic. Universal modeling
languages concentrate on graphical symbols usage laws and do not explain
symbols semantic meaning. It’s even impossible, because in each particular
case semantic may differ. As a result process semantic is hard to put into the
models and to read it and models allow different interpretations.</p>
      <p>As a result software requirements specifications and job descriptions are created in
form of a written document containing illustrative graphs and schemes.</p>
      <p>Our novelty approach is to concentrate to model semantic according to specific
model usage. We have identified 4 important model usages – informal business
description, precise business model that can be used as work instructions for people,
business process model that serves as requirement specification for IS development,
business process model that is used by IS to operate. In details – we propose to add
semantic to each symbol used in diagrams according its usage. Widely known
modeling languages, such as UML activity diagrams or BPMN can be adopted to
acquire abovementioned goals. But because of specific requirements (link with
information in existing IS, consistency checks) we decided to use specific DSL.</p>
      <p>
        Universal modeling languages specify only syntactic structure of diagram and do
not specify semantic meaning. We propose assign to activities/tasks not only unique
names but also attach informal descriptions defining semantic meaning of each
activity/task. In this case diagrams syntax can conform to UML Activity Diagrams or
BPMN or domain specific language (DSL), but semantically they are supplemented
with these informal description and shows essence of modeled system. It brings us to
specific DSL [
        <xref ref-type="bibr" rid="ref5">5</xref>
        ] that allows us to understand models in context of specific system
diagrams made with universal modeling tool.
      </p>
      <p>This technique is used in number of projects in Latvia. Result shows that proposed
approach allows minimize problems with connections between business people and IT
specialists. In this paper we explain proposed approach with small process and 4 usage
scenarios and 4 different types of models of this process.</p>
      <p>The structure of the paper is as follows. Chapter 1 contains four usage scenarios
and four different DSLs for these usage scenarios. Chapter 2 contains analysis of
properties of proposed DSLs. Chapter 3 contains our experience with usage of these
DSLs.</p>
    </sec>
    <sec id="sec-2">
      <title>1. Types of Models</title>
      <p>
        During analysis and modeling number of organizations we have developed 4 different
model usage scenarios. In this chapter we analyze these model usages and show how to
add semantics to the models. As a result we get 4 DSLs that allows to model wide
spectrum of workflow oriented systems:
1. Business process description – due to graphical illustration of the processes the
model contributes to better understanding of the nature of business processes [
        <xref ref-type="bibr" rid="ref6">6</xref>
        ].
2. Business model determines definition of a process – formalized language
determines process steps, implementation and reports, communicated among
process executors, thus creating precise descriptions of functions carried out by the
employees. Moreover, process implementation steps are closely linked to the
organizational structure – process executors and their activities under respective
normative acts. Clearly, IT specialists proposed UML only partly meets the
demands of above-mentioned non-formalized descriptions.
3. Business model serves as requirements specification for the IS development –
description of business process steps liable to IS support are included in the model
as well as all process implementation rules and algorithms.
4. Business process model creates the base of IS operations’ description [
        <xref ref-type="bibr" rid="ref7">7</xref>
        ]. Such
situation is common in recordkeeping and event-oriented systems, where
workflow is described by graphical diagrams in specialized Business Process Modeling
Language (BPML), but software is developed by generating Business Process
Execution Language (BPEL) commands from the model in BPML. Thus, we
consider the work-flow description language BPML to be domain specific
language.
      </p>
      <sec id="sec-2-1">
        <title>1.1. Model as the Process Description</title>
        <p>Usually, the initial organizational processes are missing description or descriptions are
anchored in legislative and normative acts and instructions. Frequently, these
instructions are incomplete or even contradictory and incomprehensible. Process
models are created to improve this situation. First step is to create a model that
describes process and serves as a descriptive material for organizational process. Users
of these models are usually top business managers valuing mainly comprehensibility
and simplicity of the model. Here the common principle requires intuitive guess of
model’s syntax or a minimum training. Main requirement of such models – users
without deepest understanding of modeling language must intuitively understand
process described in the model. Main components of such model are graphical
diagrams illustrating business process activities and their implementation sequence, as
well as informal textual descriptions of activities attached to diagrams. Usually, the
informal descriptions include references to normative acts, document templates and
similar documents. Graphical diagrams and informal descriptions include also other
related information such as information about process executor, deadlines and
necessary technical resources.</p>
        <p>As an example for the rest or the paper we will use registration of enterprises to
allow them to perform specific kind of entrepreneurship. A law in Latvia as in many
countries stipulates that to carry out some specific kind of entrepreneurship, for
example to provide communication services, special license must be received. Figure 1
shows the highest level model for such process. Enterprise must address application to
responsible state institution. In this case responsible institution is Public Utilities
Commission (PUC). PUC considers application in 15 days and issues license and
registers enterprise or rejects application in a case if some requirements are not
fulfilled.</p>
        <p>This simple picture can be formatted according to syntax of different modeling
language. Still the meaning of the process can be understood by reading the description
in precious paragraph. It means that main goal of modeling language is to bind process
semantics with graphical picture. Frequently semantic is described via informal
descriptions and normative acts that is not the part of the model itself. Rather process
description and normative acts are stored into separate files, different web sites or
paper documents. It is hard to use models without attaching these descriptions.</p>
        <p>The most important - theses models are easily comprehensible and well presented,
since their main objective is to provide good insight and receive management's
approval for publication. However, there is also a main deficiency – these models are
rather too abstract and fail to be completely consistent. These models usually have low
formalization level. Actually, any graphical tool, such as Microsoft Visio, that provides
unlimited choice of symbols, layouts and visualization possibilities, would be suitable
for drawing a model of such sophistication level.</p>
        <p>Representation of organizational processes by informal schemes, operational
descriptions and diagrams is a common modeling approach anchored in process
penetration and caused by requirements ISO quality management standards. These
models, which are developed for solely the purposes of quality management, contain a
certain degree of inaccuracy in order to use them for development of information
system operations or monitoring enterprise performance indicators. As a result, the
ultimate goal of quality management cannot be met, since there is a lack of objective
performance assessment and informal schemes of processes can be used only by
limited number of users.</p>
        <p>IT specialists can use such model for better understanding of processes and
industrial specifics during IS development initial phase. Informal nature of the model
constraints its use for IS development and also cannot be used as precise official
instructions.</p>
      </sec>
      <sec id="sec-2-2">
        <title>1.2. Model for the Definition of Process</title>
        <p>
          The next step of process modeling development would be to develop business process
models that define business processes and provide a precise instruction for process
execution [
          <xref ref-type="bibr" rid="ref1">1</xref>
          ]. This approach sets significantly higher and sophisticated requirements
for model's formalization, precision and consistency than in the case of models
developed solely for the purpose of process understanding. These models should have a
considerable degree of precision for employees to be able to interpret the model
information precisely and execute the process steps correctly. However, the level of
precision and formalization sufficient for employee's interpretation might not be
sufficient for the implementation of such model in information system or for the
automated execution of the model by the information system.
        </p>
        <p>In our example, the executable model contains additional information about the
process steps, activities carried out in each, including screenshots of IS forms and
reports where information must be entered and can be read during this step.
Technically this information can be seen with the corresponding model editor tool that
would allow attract the appropriate screen forms or documents to each process step.
Criteria for the models in this group - the user has access to all information necessary
to process each step in real life.</p>
        <p>Models that are used to define processes contain graphical diagrams and informal
descriptions as well as references (links) to external documents (precise references to
paragraphs in normative acts and document templates), references to functions of
external information systems (screen forms) and user instruction on filling these forms
(Figure 4), references to objects in external information system (e.g., process steps
executor in organizational structure). Model for process definition cannot bear any
inconsistency, therefore consistency verification is crucial. Conditions of consistency
check-ups should be defined in the same modeling language and check-up should be
performed by the modeling tool. It should be emphasized that consistency conditions
are domain specific, which means – theses conditions are specific to organizational and
business requirements of a particular organization.</p>
        <p>Main users of model are employees, which use the model as an instruction for
everyday execution of business processes. Above mentioned model does not contain
precise data structures, therefore it cannot be automatically executed and only partially
contains information necessary for IS requirements specification.</p>
      </sec>
      <sec id="sec-2-3">
        <title>1.3. Model as IS Requirements Specification</title>
        <p>Model described in the previous sections were used to define business people view of
processes taking place in organization. In abovementioned model business processes
are described maximally precise from business point of view so that employees can
accomplish them. Business people and also IT people are able to understand language
used in such models. However, these definitions and descriptions are not sufficient in
order to describe operations of information system – these models must be
complemented with data structures and other information required to develop IS. Such
models we will call IS diagrams.</p>
        <p>Usually IS diagrams are created by the information system analysts. IS diagrams
contain descriptions of software screen forms, reports and automated tasks –
algorithms. IS diagrams describe conceptual data level, which can be detailed further to
the level of physical data model. IS diagrams continue detailing the business process
steps from the point where business process diagrams have reached their constraints.
For example we take process step “Make entry in register” form previous model in
Figure 3 and detail it. We specify this step as two steps and add IS menu items needed
to perform this step (Figure 5).</p>
        <sec id="sec-2-3-1">
          <title>IS diagrams have several ways of applicability:</title>
          <p>•
•
•
•</p>
          <p>IS diagrams serve as a basis for development of software functional
requirements containing data input, data output, data processing and data
verification.</p>
          <p>IS diagrams determine, which business process fragments should be
implemented in information system and became fully automated and which
should remain manual even after information system is introduced. For
example, if a business process foresees verification of a signature on
submitted document then the employee will perform the verification manually.
IS can only verify input values and information consistency.</p>
          <p>IS diagrams allows to make data usability analysis. For, example, if we would
like to change processing logic of data object we can create view where we
can see all IS activities that use or influence these data. A different view can
be created in order to monitor status change of an object in order to indicate
activities related to object status and its changes.</p>
          <p>IS diagrams can serve as a basis for creation of test cases for software testing.
Use of both, business process model and IS diagrams, can easily solve the
problem of software testing completeness - business process model can be
used as testing model.</p>
          <p>In summary, if business requirements are gradually and successfully developed
into IS requirements, organization obtains a very precise description of organization's
information system and IT specialists obtain functional requirements for the system
implementation. By using and maintaining business process and IS models in a long
term organization obtains significant information on impacts and costs of changing or
adding new business functions to organization.</p>
        </sec>
      </sec>
      <sec id="sec-2-4">
        <title>1.4. Model as IS Component</title>
        <p>Current trends in information system technological development can be characterized
by less technical coding, more precise specification resulting in software development
"without programming". Modeling is the integral part of this process.</p>
        <p>IT trend, where modeling builds the foundation to systems operations, is
represented in workflow-oriented systems. A workflow indicating process steps
activities is developed according to organization's business process. In a workflow
significant role is played by executor – user, which has an assigned role in
organizational structure. Process steps perform data processing of involved object.
Work-flow provides possibility to store data throughout the object's life-cycle. Such
concepts and terms as task deadlines, actual status of object, possibility to create
documents automatically during workflow execution are very significant in work-flow
oriented systems.</p>
        <p>There are two specific tasks for models, first, it should be rather simple and
comprehensible for employee to understand workflow, second, it should be
sophisticated enough in order to contain all information significant for operation of
information system. In principal, these are two different views of the same workflow
diagram.</p>
        <p>Which would be the most appropriate language for creation of workflow
description? One possibility is use UML or BPMN. Unfortunately, in the UML and
BPMN models because of their universality is difficult to incorporate domain
semantics. In addition, the UML and BPMN are overloaded with many technical
features, which are needed for the development of IS in specific cases, but often are
redundant for a particular system and is difficult to understand by the business people.
Frequently acceptable option is to use a DSL, which allows you to build models for
both – business needs and for the IS implementation.</p>
        <p>
          Authors of this article have developed domain specific language BILINGVA [
          <xref ref-type="bibr" rid="ref7">7</xref>
          ],
with the purpose to describe workflows. Diagrams illustrate object statuses, process
steps (events) and sequences of process steps. The language provides possibility to
describe also decisions – branching points in processes, define deadlines and execute
processes parallel. After a model is created the information stored in diagrams is
transported to the information system repository and system’s configuration tables. By
operating information system its engine interprets information obtained from diagrams
automatically, thus ensuring system functionality defined in the model.
        </p>
        <p>Figure 5 contains executable enterprise registration process. Unlike diagram from
previous chapter, all activities in this model can be executed by IS. The model does not
contain objects that are not covered by IS such as “Enterprise” and “Enterprises
register”. The model also does not contain manual operation such as document review.
In BILNGVA syntax arrows are activity/task but boxes are object states. Parallel
execution also can be shown. After application, registration, enforcement activity has
been transferred to the Accounting Division, operating as sub process in parallel with
the main process, and shall be executed by an expert of Electronic Division.</p>
        <p>Time limit of the process is represented as two events – “begin time control” and
“test time control”. In event “begin time control” event “Time to register – 15 days” is
stored. IS form shown to user in a case time control rule is infringed is implemented
into model interpreter and has not specified in the model. This helps to create
userfriendly interface for execution of concrete activities. Such user friendliness is hard to
achieve in a case if universal solutions are used.</p>
        <p>
          Diagram in Figure 6 is a user-oriented view, where the notation proposed in [
          <xref ref-type="bibr" rid="ref7">7</xref>
          ] is
used. To ensure correct work of system process modeler completes attribute set for
each process step that is necessary for correct operation of system. For example each
step is characterized by the following parameters – does step can make document, if so,
than document template must be specified, does some document can be uploaded to
system during the step, does specific checklist is fulfilled during the step, step
executors roles – manager, responsible employee or anyone. Does responsible
employee can be set in this step? In which conditions step can be executed? Step status
attribute set describing which departments see this status, status number - to sort
statuses in user interface in logical order, status group. Time limit is additionally
characterized by following attributes: is time limit in calendar days or working days, is
time limit automatically or manually counted, how many days before deadline system
issues warning about deadline, which departments are bound to the deadline.
        </p>
        <p>Returning to applicability aspect, a detailed workflow diagram provides
framework for system's operations, a special view serves as a user's manual, which can
be complemented by explanatory documents referenced to certain steps.</p>
        <p>The future perspective includes task – to implement graphical representation of
process step execution in business process diagram visible to system user. For example,
by providing simplified process model on organization's homepage the customers could
track and follow execution status of provided service.</p>
      </sec>
    </sec>
    <sec id="sec-3">
      <title>2. Components of Model</title>
      <sec id="sec-3-1">
        <title>2.1. Models Attributes</title>
        <p>Each of four models described above have their own applications and users.
Consequently, these models differ in the semantics of elements and attributes of the
elements. This fact is presented in Table 1 where some attributes of the process steps of
various models are given.</p>
        <sec id="sec-3-1-1">
          <title>Documents bound to the model can differ according to the model type:</title>
          <p>•</p>
          <p>Model for purpose of process – process regulating documents are attached as
external files. Usually documents are attached to the whole process not to
specific process step.
•
•</p>
          <p>Model for definition of process –regulatory documents preferable attach to
specific process step. IS screenshots, document templates, user manuals and
job descriptions are attached to the process steps.</p>
          <p>Model for IS requirements specification –specific chapter regulating process
step preferable attach to process step. IS forms that must be fulfilled in this
step, description of used data (data model), consistency checks, error
messages and other information necessary for requirements specification are
attached to the process step.
•</p>
          <p>Model as IS component – document templates used in this step to generate
documents are attached to the process step.</p>
          <p>Other attributes can be added to the model according to the domain specific.</p>
        </sec>
      </sec>
      <sec id="sec-3-2">
        <title>2.2. Model Views</title>
        <p>Model describes the operation of a single object - the organization, its processes and
their implementation. Our approach offers to build a single model with common
information and add different views to model for different users and different
applications. Various pieces of information from the model (different views of the
model) are necessary for business representatives, customers, quality service, IT
professionals, and so on. The above mentioned applications show how wide are usage
of models. We will mention number of other applications in this chapter, which
requires their own solutions.</p>
        <p>One way to use model is different views for different applications. Example of
view is series of activities performed to accomplish specific service (for example,
registration of enterprises to allow them to perform specific kind of entrepreneurship).
Another example of a view - the process steps that specific position is interested in - for
example, Manager view of the processes, which include not only the steps Manager
performs, but also actions that are performed before and after him. Another example of
view - the quality system view of the organization's operations, which includes the
entire range of activities, but rather superficial, omitting a number of components.</p>
        <p>We may say that Manager is interested in some of the process execution to the
smallest detail, but in some other process execution is actually not interested at all. By
contrast, the quality department is interested in all the processes, but to the highest top
level without going into detail.</p>
        <p>It is unacceptable that implementation of the above mentioned views would be
generated in a separate model, unrelated to the common organizational model. In order
to maintain consistency each such view should be implemented as a subset of the
overall model. Currently, such views can be created manually and consistency checks
can be built into tool. At present, an unresolved problem is the view definition and
automated creation from the integrated model.</p>
        <p>Moreover, the model can be applied in a further way – service diagrams can be
developed based on functionality model. Service diagrams illustrate certain process
steps to be carried out in order to provide a particular service to customers. For
example, a help-desk inspector at one stop agency provides services to a number
customer with various demands. It is impossible for inspector to precisely know by
heart every process step of a number of services applicable for each customer case.
Therefore the inspector uses process model that includes precise instruction for each
service case.</p>
        <p>A different reporting possibility is a further way of model applicability, for
example, reference of process steps to certain legislative acts. In case of amendments of
legislative acts these references easily indicate process steps that should be changed.
UML and partly also BPMN can be used as a language to meet the requirements of
such models for these languages provide precise definition for process execution.
However, missing possibility of adding domain specific information to model makes
use of theses languages problematic, for example, would it be possible to easily added
references to paragraphs of normative acts if a process step is defined in UML or
BPMN? Moreover, since the consistencies usually are domain specific, the use of
universal modeling languages makes keeping of model consistencies problematic. The
elimination of this problem would require development of specific tools in order to
ensure consistency in model created by universal languages.</p>
      </sec>
    </sec>
    <sec id="sec-4">
      <title>3. Real Life Experience</title>
      <p>
        Authors has several years of experience in the use of DSL in seven organizations in all
four analyzed above applications. Usually each case consisted of three steps. The first
step was DSL definition, typically through business process modeling together with the
company's business professionals. Usually new DSL was based on some already
existing DSL and a new definition seemed to add extra attributes and attach
domainspecific input / output documents. DSL definition virtually at the same time meant the
definition of modeling tool. That was made possible by using a DSL tool building
platform [
        <xref ref-type="bibr" rid="ref10 ref8 ref9">8, 9, 10</xref>
        ]. In the third step, the business model was developed using newly
created DSL and modeling tool. This model was then transferred to a business-specific
application. Accumulated experience has shown this approach strength:
Each new DSL definition reveals new requirements for organizational model,
such as model’s elements linkage to the sector-specific document flow.
Identified requirements can be incorporated into DSL and model.
      </p>
      <p>General purpose modeling language cannot describe all the sector specific
requirements, for example, it’s hard to access to non-formal process
descriptions stored in the enterprise IS. Similarly, if universal modeling
language is used then it is hard to link document samples available in
company IS or website to the process descriptions.</p>
      <p>It was especially surprising to discover very positive attitude of users towards
graphical specifications which users preferred to use as information system
exploitation manual for it contained more precise information in a more
comprehensible way, thus replacing old-fashioned user manuals.</p>
      <p>Modeling technology is a new type of activity for business people. It takes time and the
gradual acquisition of modeling techniques. At first, business people are able to create
informal models (informal process description models) and only after some time they
are able to move to more formal models (formal models of process definition). In our
case adoption of modeling technique and models in different organizations took 2-4
years.</p>
    </sec>
    <sec id="sec-5">
      <title>4. Conclusions and Proposals</title>
      <p>The experience of DSL’s use allows make such main conclusions. Strength of DSL:</p>
      <p>DSL allows create single unified model within organization for all four
different usage scenarios mentioned in this paper and allows different views to
the model.</p>
      <p>Users can easily understand meaning of models and use models if business
process semantic is bind to model objects. Syntax specific usually does no
matter so much.</p>
      <p>DSL building is one of easiest way to bind semantic of specific domain to
model.</p>
      <p>
        DSL definition and modeling tool definition platform plays one of mayor role
for DSL usage. It is practically impossible to implement in real life
methodology we describe without such platform. Platform we use [
        <xref ref-type="bibr" rid="ref10 ref8 ref9">8,9,10</xref>
        ]
ensures DSL definition and modeling tool definition in relatively small time
frame (approximately 3 month).
      </p>
      <p>Survey shows that end users definitely prefer graphical diagrams instead of
traditional text documents. Experience shows that one graphical diagram can
contain the same amount of information as 10 pages of textual instructions.
Proposed process modeling methodology is examined in workflow type
systems. For other types of systems the situation may differ.</p>
      <p>Definition of DSL and development of modeling tools requires involving high
qualification specialists.</p>
      <p>Enterprise specific DSL development and business process definition is
individual as enterprise specific IS development. Clients must support not
broadly recognized DSL technologies, in order to develop and maintain DSL,
modeling tool and company models.</p>
      <p>DSL defined for needs of one company is hard or impossible use in other
company even if the companies’ profiles are very similar. Experience so far
demonstrates, that DSL for each enterprise contains nuances specific for each
enterprise. Previously developed DSLs can be used in the very beginning of
modeling and help to recognize specific of new domain.</p>
      <p>
        We agree to the supporters of MDSD about growing role of models in IS
development process [
        <xref ref-type="bibr" rid="ref11">11</xref>
        ] and we plan to study modeling and usage of DSL as one of
method for company business process arranging and IS developing more closely in
near future.
      </p>
    </sec>
    <sec id="sec-6">
      <title>Acknowledgement</title>
      <sec id="sec-6-1">
        <title>The work is supported by a European 2009/0216/1DP/1.1.1.2.0/09 /APIA/VIAA/044.</title>
      </sec>
      <sec id="sec-6-2">
        <title>Social</title>
      </sec>
      <sec id="sec-6-3">
        <title>Fund</title>
      </sec>
      <sec id="sec-6-4">
        <title>Project No.</title>
      </sec>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <given-names>T.</given-names>
            <surname>Stahl</surname>
          </string-name>
          and
          <string-name>
            <given-names>M.</given-names>
            <surname>Markus</surname>
          </string-name>
          ,
          <string-name>
            <surname>Model-Driven Software</surname>
            <given-names>Development</given-names>
          </string-name>
          , John Wiley &amp; Sons,
          <year>2006</year>
          .
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2]
          <string-name>
            <given-names>MDA</given-names>
            <surname>Guide</surname>
          </string-name>
          <article-title>Version 1.0.1</article-title>
          . OMG. Available from: http://www.omg.org/docs/omg/03-06-01.pdf.
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          [3]
          <string-name>
            <given-names>BPMN</given-names>
            <surname>- Business Process</surname>
          </string-name>
          Model and Notation. OMG. Available from: http://www.bpmn.org.
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          [4]
          <string-name>
            <given-names>UML</given-names>
            <surname>- Unified Modeling Language. OMG</surname>
          </string-name>
          . Available from: http://www.uml.org.
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          [5]
          <string-name>
            <given-names>J.</given-names>
            <surname>Bicevskis</surname>
          </string-name>
          ,
          <string-name>
            <given-names>J.</given-names>
            <surname>Cerina-Berzina</surname>
          </string-name>
          ,
          <string-name>
            <given-names>G.</given-names>
            <surname>Karnitis</surname>
          </string-name>
          ,
          <string-name>
            <given-names>L.</given-names>
            <surname>Lace</surname>
          </string-name>
          ,
          <string-name>
            <surname>I. Medvedis</surname>
          </string-name>
          , and
          <string-name>
            <given-names>S.</given-names>
            <surname>Nesterovs</surname>
          </string-name>
          ,
          <article-title>Practitioners view on domain specific business process modeling</article-title>
          . In: J.
          <string-name>
            <surname>Barzdins</surname>
          </string-name>
          , M. Kirikova, editors,
          <source>Databases and Information Systems VI, Selected papers from 9th International Baltic Conference DB&amp;IS</source>
          <year>2010</year>
          , IOS Press,
          <year>2011</year>
          ,
          <fpage>169</fpage>
          -
          <lpage>182</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          [6]
          <string-name>
            <given-names>A.</given-names>
            <surname>Schleicher</surname>
          </string-name>
          ,
          <article-title>High-level modelling of development processes</article-title>
          . In: B.
          <string-name>
            <surname>Scholz-Reiter</surname>
            ,
            <given-names>H. D.</given-names>
          </string-name>
          <string-name>
            <surname>Stahlmann</surname>
            ,
            <given-names>A</given-names>
          </string-name>
          . Nethe, editors,
          <source>Process Modelling</source>
          . Springer,
          <year>1999</year>
          ,
          <fpage>57</fpage>
          -
          <lpage>73</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          [7]
          <string-name>
            <given-names>J.</given-names>
            <surname>Cerina-Berzina</surname>
          </string-name>
          ,
          <string-name>
            <given-names>J.</given-names>
            <surname>Bicevskis</surname>
          </string-name>
          , and G. Karnitis,
          <article-title>Information systems development based on visual domain specific language BiLingva</article-title>
          . In: T. Szmuc,
          <string-name>
            <given-names>M.</given-names>
            <surname>Szpyrka</surname>
          </string-name>
          , J. Zendulka, editors,
          <source>Advances in Software Engineering Techniques, Lecture Notes in Computer Science</source>
          <volume>7054</volume>
          (
          <year>2011</year>
          ), Springer, Berlin,
          <fpage>124</fpage>
          -
          <lpage>135</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          [8]
          <string-name>
            <given-names>J.</given-names>
            <surname>Barzdins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>A.</given-names>
            <surname>Zarins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>K.</given-names>
            <surname>Cerans</surname>
          </string-name>
          ,
          <string-name>
            <given-names>A.</given-names>
            <surname>Kalnins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>E.</given-names>
            <surname>Rencis</surname>
          </string-name>
          ,
          <string-name>
            <given-names>L.</given-names>
            <surname>Lace</surname>
          </string-name>
          ,
          <string-name>
            <given-names>R.</given-names>
            <surname>Liepins</surname>
          </string-name>
          ,
          <article-title>and</article-title>
          <string-name>
            <given-names>A.</given-names>
            <surname>Sprogis</surname>
          </string-name>
          ,
          <article-title>GrTP: transformation based graphical tool building platform</article-title>
          .
          <source>In: Proceedings of the MoDELS 2007 Workshop on Model Driven Development of Advanced User Interfaces</source>
          , Nashville, Tennessee, USA, October
          <volume>1</volume>
          ,
          <year>2007</year>
          . Available from: http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-
          <volume>297</volume>
          /.
        </mixed-citation>
      </ref>
      <ref id="ref9">
        <mixed-citation>
          [9]
          <string-name>
            <given-names>J.</given-names>
            <surname>Barzdins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>S.</given-names>
            <surname>Kozlovics</surname>
          </string-name>
          , and
          <string-name>
            <surname>E. Rencis,</surname>
          </string-name>
          <article-title>The transformation-driven architecture</article-title>
          .
          <source>In: Proceedings of DSM'08 Workshop of OOPSLA</source>
          <year>2008</year>
          ,
          <article-title>Nashville</article-title>
          , USA,
          <year>2008</year>
          ,
          <fpage>60</fpage>
          -
          <lpage>63</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref10">
        <mixed-citation>
          [10]
          <string-name>
            <given-names>J.</given-names>
            <surname>Barzdins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>A.</given-names>
            <surname>Zarins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>K.</given-names>
            <surname>Cerans</surname>
          </string-name>
          ,
          <string-name>
            <given-names>M.</given-names>
            <surname>Grasmanis</surname>
          </string-name>
          ,
          <string-name>
            <given-names>A.</given-names>
            <surname>Kalnins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>E.</given-names>
            <surname>Rencis</surname>
          </string-name>
          ,
          <string-name>
            <given-names>L.</given-names>
            <surname>Lace</surname>
          </string-name>
          ,
          <string-name>
            <given-names>R.</given-names>
            <surname>Liepins</surname>
          </string-name>
          ,
          <string-name>
            <given-names>A.</given-names>
            <surname>Sprogis</surname>
          </string-name>
          ,
          <article-title>and</article-title>
          <string-name>
            <given-names>A.</given-names>
            <surname>Zarins</surname>
          </string-name>
          ,
          <article-title>Domain specific languages for business process managment: a case study</article-title>
          .
          <source>In: Proceedings of DSM'09 Workshop of OOPSLA</source>
          <year>2009</year>
          ,
          <article-title>Orlando</article-title>
          , USA,
          <year>2009</year>
          . Available from: http://www.dsmforum.org/events/DSM09/.
        </mixed-citation>
      </ref>
      <ref id="ref11">
        <mixed-citation>
          [11]
          <string-name>
            <given-names>L.</given-names>
            <surname>Cao</surname>
          </string-name>
          ,
          <string-name>
            <given-names>B.</given-names>
            <surname>Ramesh</surname>
          </string-name>
          , and
          <string-name>
            <given-names>M.</given-names>
            <surname>Rossi</surname>
          </string-name>
          ,
          <article-title>Are domain-specific models easier to maintain than UML models</article-title>
          ?
          <source>IEEE Software 26(4)</source>
          (
          <year>2009</year>
          ),
          <fpage>19</fpage>
          -
          <lpage>21</lpage>
          .
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>