=Paper=
{{Paper
|id=Vol-2915/paper13
|storemode=property
|title=Knowledge-Based UML Use Case and UML Activity Models Generation from Enterprise Model. School of Languages Case Study
|pdfUrl=https://ceur-ws.org/Vol-2915/paper13.pdf
|volume=Vol-2915
|authors=Ilona Veitaitė,Audrius Lopata
|dblpUrl=https://dblp.org/rec/conf/ivus/VeitaiteL21
}}
==Knowledge-Based UML Use Case and UML Activity Models Generation from Enterprise Model. School of Languages Case Study==
https://ceur-ws.org/Vol-2915/paper13.pdf
Knowledge-Based UML Use Case and UML Activity Models
Generation from Enterprise Model.
School of Languages Case Study
Ilona Veitaitėa and Audrius Lopatab
a
Vilnius University, Kaunas Faculty, Muitinės str. 8, Kaunas, LT-44280, Lithuania
b
Kaunas University of Technology, Studentų str. 50, Kaunas, LT-51368, Lithuania
Abstract
The main purpose of this paper is to present knowledge-based Enterprise model (EM)
advantages as data repository for UML models generation. UML models can be generated from
Enterprise Model, whose main requirement is that gathered data in this model would be verified
and validated by analyst. UML models generation process is implemented by using particular
transformation algorithms presented in previous researches. In this paper generation process
from EM is represented by the School of Languages case study. By using problem domain data
UML Use Case and UML Activity models are generated. Generation results help to define
advantages of Enterprise Model usage as storage for problem domain data.
Keywords 1
Enterprise Model, UML, IS Engineering, Activity, Use Case, Knowledge-based.
1. Introduction
Today’s IT professionals such as analysts, designers, developers still face with challenges of IS
engineering process. First phases of IS development life cycle is also quite difficult, especially,
enterprise modelling, when all gathered data is specified and prepared for IS design phase. According
the efforts put in this phase success of final IS depends. There are many various types, standards of
models used in design phase and very great impact for IS creation success has Enterprise model chosen
for this process [1][2]. UML is one of mostly used standard for IS design among professionals of IS
engineering field. According to UML models prepared in design phase IS code may be generated,
different design patterns applied, impact and complexity analysis accomplished [1][6][8]. All
advantages of UML models usage may be fulfilled only if data used for their design is verified and
validated. This fulfillment can be ensured by Enterprise model. Enterprise model - knowledge-based
repository, where problem domain data of enough quality is stored. By using transformation algorithms
all UML models can be generated from Enterprise Model [3][5][7][9][10]. Structure of particular
Enterprise Meta-Model (EMM) and Enterprise Model as the background for the research in this paper
is presented almost two decades ago. Previous researches in this field are designated to prove that
Enterprise meta-model and Enterprise Model is enough for different types of models generation in IS
design phase [10][11][12][13][14].
2. Description of Knowledge-Based Enterprise Model
EMM is formally defined Enterprise Model structure, which consists of a formalized EM in line
with the general principles of control theory. EM is the main source of the necessary knowledge of the
particular business domain for IS engineering and IS re-engineering processes (Figure 1) [3][4].
IVUS2021: Information Society and University Studies 2021, April 23, 2021, Kaunas, Lithuania
EMAIL: ilona.veitaite@knf.vu.lt (A. 1); audrius.lopata@ktu.lt (A. 2)
©️ 2021 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
Figure 1: Enterprise Meta-Model class diagram [3][4][10]
EM class model has twenty-three classes. Essential classes are Process, Function and Actor. Class
Process, Function, Actor and Objective can have an internal hierarchical structure. These relationships
is presented as aggregation relationship. Class Process is linked with the class MaterialFlow as
aggregation relationship. Class MaterialFlow is linked with the classes MaterialInputFlow and
MaterialOutputFlow as generalization relationship. Class Process is linked with Classes Function,
Actor and Event as association relationship. Class Function is linked with classes InformationFlow,
InformationActivity, Interpretation, InformationProcessing and Realization as aggregation relationship.
These relationships define the internal composition of the Class Function. Class InformationFlow is
linked with ProcessOutputAtributes, ProcessInputAtributes, IPInputAttributes and IPOutputAttributs
as generalization relationship. Class InformationActivity is linked with Interpretation,
InformationProcessing and Realization as generalization relationship. Class Function linked with
classes Actor, Objective and BusinessRule as association relationship. Class BusinessRule is linked
with Interpretation Rule, Realization Rule, InformationProcessing Rule as generalization relationship.
Class Actor is linked with Function Actor and Process Actor as generalization relationship
[3][4][5][11].
3. Transformation Algorithms from Knowledge-Based Enterprise Model
Each of static and dynamic UML models can be generated through transformation algorithm and
each of UML models has separate transformation algorithm. These transformation algorithms are
defined in previous researches. Main focus of researches is designated for generation behavioural or
dynamic UML models, because their complexity and variability [10][11][12][13][14][15].
3.1. UML Use Case Model Transformation Algorithm
UML Use Case model generation from Enterprise Model transformation algorithm starts from initial
element. In this generation process Actor/Subject is initial element, after actor/subject element is
generated, Use Case element is selected and generates, then Include, Extend and Association
relationships elements are selected and generated. Transformation algorithm is illustrated by following
steps [12]:
Step 1: The initial element Actor from Enterprise Model for UML Use Case model
generation is selected.
Step 2: If Actor element is initial element of UML Use Case model, then Actor element is
generated, else Subject element is generated.
Step 3: Process element from Enterprise model, which is related with the initial Actor
element is selected.
Step 4: If Process element is Use Case element related to Actor/Subject, then Use Case
element is generated, else Function element is selected.
Step 5: Function element is generated as Use Case element.
Step 6: Business Rule element as link of Actor/Subject element from Enterprise Model
which is related with the Process/Function element is selected.
Step 7: If Business Rule element is UML Use Case model’s simple Association element and
serves as link between Actor/Subject and Process/Function elements then Association is
generated from Enterprise model, else if it is Extend element, then Extend element is
generated from Enterprise model, else Include element is generated from Enterprise model.
Step 8: There is checking if there are more Business Rules in Enterprise Model related to
UML Use Case model. In case, there are, algorithm goes back to step 3.
Step 9: UML Use Case elements Actor/Subject and Process/Function are linked according
to Business Rules.
Step 10: UML Information flow element Actor/Subject is updated.
Step 11: There is checking if there are more Actors/Subject elements in Enterprise Model
related to UML Use Case model. In case, there are, algorithm goes back to step 1.
Step 12: Else all UML Use Case model elements and links are generated from Enterprise
Model.
3.2. UML Activity Model Transformation Algorithm
UML Activity model describes how activities are coordinated to provide a service. UML Activity
model from EM transformation algorithm is described by following steps [14]:
Step 1: Particular UML model for generation from EM process is identified and selected.
Step 2: If the particular UML model for generation from EM process is selected then
algorithm process is continued, else the particular UML model for generation from EM
process must be selected.
Step 3: First element from EM is selected for UML model, identified previously, generation
process.
Step 4: If the selected EM element is initial UML model element, then initial element is
generated, else the other EM element must be selected (the selected element must be initial
element).
Step 5: The element related to the initial element is selected from Enterprise model.
Step 6: The element related to the initial element is generated as UML model element.
Step 7: The element related to the previous element is selected from Enterprise model.
Step 8: The element related to the previous element is generated as UML model element.
Step 9: If there are more related elements, then they are selected from EM and generated as
UML model elements one by one, else the link element is selected from Enterprise model.
Step 10: The link element is generated as UML model element.
Step 11: If there are more links, then they are selected from EM and generated as UML
model elements one by one, else the Business Rule element is selected from Enterprise
model.
Step 12: The Business Rule element is generated as UML model element.
Step 13: If there are more Business Rules, then they are selected from EM and generated as
UML model elements one by one, else the generated UML model is updated with all
elements, links and constraints.
Step 14: Generation process is finished.
4. School of Languages Case Study
The School of languages management process may be defined as courses management system that
is used to manage language courses, timetable, lecturers and participants of the courses.
The School of languages offers a list of language courses related with different level of language
knowledge and diverse study methods, which improve speaking, reading, listening, speaking and
grammar skills. Each language course is made up of set of topics related with skills dedicated for
improvement. The School of languages publish and maintain a timetable of different language courses
and appoints lecturers every year.
Lecturers in the school of languages are appointed courses to teach according to the language
knowledge level and their availability. There is a course administrators in the school of languages to
manage the courses including course content, courses appointments to lecturers and preparation of the
course timetable. Participants of language courses may review suggested courses, get information about
lecturers and check the timetable of the courses.
The School of Languages purpose is to use the courses management information system to improve
their services by managing courses, timetable, lecturers and participants.
4.1. School of Languages UML Use Case Model
The main concept of UML Use Case model is that it assists to design a system from the end user's
perspective, in this case, end users are Administrator, Lecturer and Participant.
Figure 2: UML Use Case Model.
Figure 2 presents generated UML Use Case model from Enterprise model, where uses cases of each
actor are displayed. Descriptions of these generated UML Use Case Model elements are presented in
the Table 1.
Table 1
Generated UML Use Case Model elements from EM with descriptions [10] [12]
EM UML Use Case Model elements Description
elements
Actor Actor: Administrator Administrator has full access to all
use cases implemented in School of
languages IS.
Actor: Lecturer Lecturer can check appointed to him
language course, review list of
participants of the appointed
language course and review his
timetable.
Actor: Participant Participant is a student of chosen
language courses, he can review list
of courses, review information about
lecturers and also review timetable
of chosen courses.
Business Association Relationship that links all actors to
Rules the use cases they may access
Extend Relationship that links use case,
which may be done together with
another use case. During the course
management process, topics may be
added
Process / Use Case: Manage language course Use cases which are accessible to the
Function Use Case: Manage topics for language course actors.
Use Case: Appoint language course to lecturer
Use Case: Review course participants
Use Case: Review language courses timetable
Use Case: Manage lecturer information
Use Case: Review lecturers
Use Case: Review language course
Use Case: Assign participant to course
Use Case: Manage participant information
4.2. School of Languages UML Activity Models
UML Activity models present coordination of activities regarding process participants:
Administrator, Lecturer and Participant. There can be more UML Activity model generated and at
different levels of abstraction, but in this paper three possible models are presented, only to present this
possibility.
4.2.1. UML Activity Model. Actor: Administrator
UML Activity model for language course management performed by Administrator is generated
from Enterprise model.
Figure 3: UML Activity Model. Actors: Administrator
Figure 3 presents generated UML Activity model from Enterprise model, where activities of actor,
in this case, Administrator are displayed. Descriptions of these generated UML Activity Model
elements are presented in the Table 2.
.
Table 2
Generated UML Activity Model elements from EM with descriptions [10] [14]
EM elements UML Activity Model elements Description
Actors Swimlane: Administrator Administrator starts the process and
performs all presented activities.
Process / Activity: Manage language course All activities performed by
Function Activity: Create language course Administrator of the particular
Activity: Create topic of the language process.
course
Activity: Appoint lecturer
Activity: Prepare timetable of
language course
Activity: Modify language course
Activity: Remove language course
Activity: Language course updated
Information Object Flows All flows between activities, show
Flow sequence of activities flow.
Business Control Node: Join Node Joins three different activities related
Rules with language course management.
Control node: Initial Node Starts the process.
Control node: Decision Node Two decision nodes: checks course
status; another allows modify or
remove the course.
Control node: Final Node Finishes process.
4.2.2. UML Activity Model. Actors: Administrator and Lecturer
UML Activity model for lecturer information management performed by Administrator and
Lecturer is generated from Enterprise model.
Figure 4: UML Activity Model. Actors: Administrator and Lecturer
Figure 4 presents generated UML Activity model from Enterprise model, where activities of actors,
in this case, Administrator and Lecturer are displayed. Descriptions of these generated UML Activity
Model elements are presented in the Table 3.
Table 3
Generated UML Activity Model elements from EM with descriptions [10] [14]
EM UML Activity Model elements Description
elements
Actors Swimlane: Administrator Administrator starts the process for
lecturer information management;
appoints lecturer to the language
course, prepares timetable and assigns
participants to the language course.
Swimlane: Lecturer Lecturer may review language courses
timetable, confirms appointed language
courses, confirms timetable and may
review course participants.
Process / Activity: Manage lecturer information All activities performed by Actors of
Function Activity: Appoint lecturer the particular process.
Activity: Prepare timetable of language
courses
Activity: Review language course
timetable
Activity: Confirm appointed language
course
Activity: Confirm language courses
timetable
Activity: Assign participants to the
course
Activity: Review course participants
Information Object Flows All flows between activities, show
Flow sequence of activities flow.
Business Control node: Initial Node Starts the process.
Rules Control node: Decision Nodes Two decision nodes: one confirms
appointed course; another confirms
course timetable.
Control node: Final Node Finishes process.
4.2.3. UML Activity Model. Actors: Administrator and Participant
UML Activity model for participant information management performed by Administrator and
Participant is generated from Enterprise model.
Figure 5: UML Activity Model. Actors: Administrator and Participant
Figure 5 presents generated UML Activity model from Enterprise model, where activities of actors,
in this case, Administrator and Participant are displayed. Descriptions of these generated UML Activity
Model elements are presented in the Table 4.
Table 4
Generated UML Activity Model elements from EM with descriptions [10] [14]
EM elements UML Activity Model elements Description
Actors Swimlane: Administrator Administrator starts the process of
participant information
management; prepares timetable of
the courses and assigns participants
to the course.
Swimlane: Participant Participant may review the courses
descriptions, review lecturers
responsible for the course, review
timetable and confirm it.
Process / Activity: Manage participant information All activities performed by Actors of
Function Activity: Prepare timetable of language the particular process.
courses
Activity: Review language course
Activity: Review lecturers
Activity: Review language courses
timetable
Activity: Confirm language courses
timetable
Activity: Assign participants to the course
Information Object Flows All flows between activities, show
Flow sequence of activities flow.
Business Control node: Initial Node Starts the process.
Rules Control node: Decision Node One decision node: confirms course
timetable
Control node: Final Node Finishes process.
5. Results
Main activities of the analyst responsibility may be described as: analysis of enterprise situation,
identification of opportunities for improvements, design information system which will add value to
the enterprise. The analyst gathers problem domain information, identifies all requirements, searches
for suitable meta-model, verifies and validates data and starts implement IS by designing UML or other
models. There is always risk of new problem, information or requirements appearance, which cause
difficult design models updating and improving process. In this case, duration of IS developing process
increase, and number of errors grows. Using Enterprise Model as the prime problem domain knowledge
repository in IS engineering process ensures correctness and quality of generated IS design models after
any possible problem domain data update (Table 5).
Table 5
Comparison of IS Analyst’s activities by criteria
Criteria IS Analyst Enterprise Knowledge-Based UML
Dynamic Models Generation Method
Gathering problem domain data Yes Yes
Requirements identification Yes Yes
Data preparation for modelling Yes Yes
Project models design Yes No
Problem domain data update Yes Yes
Project model design improvement Yes No
Increased IS development process duration Yes No
Increased number of errors Yes No
Increased number of errors Yes No
By using Enterprise Model in IS engineering process the analysts enters all gathered problem domain
data into EM. Problem domain knowledge stored in EM is used for UML model generation through
transformation algorithms. After any possible new data upload to the EM, it is re-used and new UML
models based on improved data are generated. There is no need for the analyst to re-do entire process
of models design.
6. Conclusions
The first part of the paper presents the Enterprise model structure and description of UML Use Case
and UML Activity models transformation algorithms for generation from EM depicted by steps.
The next part presents School of Languages case study, which data is stored in knowledge-based
Enterprise Model and is used in two types UML models generation process. There are presented UML
Use Case model and three UML Activity models of different Actors perspective generated from
Enterprise Model. These generated models confirms, that data stored in Enterprise model is enough for
generation process.
Final part describes comparison by certain criteria of information systems analyst’s activities during
IS development process with and without usage of Enterprise knowledge-based UML models
generation method.
7. References
[1] Dunkel J., Bruns R. (2007) Model-Driven Architecture for Mobile Applications. Proceedings
of the 10th Inter-national Conference on Business Information Systems (BIS), Vol. 4439/2007,
pp. 464–477.
[2] Eichelberger H., Eldogan, Y., Schmid K. A (2011) Comprehensive Analysis of UML Tools,
their Capabilities and Compliance. Software Systems Engineering. Universität Hildesheim.
versio 2.0
[3] Gudas S. (2009) Architecture of Knowledge-Based Enterprise Management Systems: a Control
View. Proceedings of the 13th world multiconference on systemics, cybernetics and informatics
(WMSCI2009),) July 10 – 13, Orlando, Florida, USA, Vol. III, p.161-266 ISBN -10: 1-
9934272-61-2 (Volume III).ISBN -13: 978-1-9934272-61-9
[4] Gudas S., (2012) Informacijos sistemų inžinerijos teorijos pagrindai/ Fundamentals of
Information Systems Engineering Theory. (Lithuanian)Vilnius University. ISBN 978-609-
459-075-7
[5] Jacobson, I., Rumbaugh, J., Booch, G. (2005) Unified Modeling Language User Guide, The
Second Edition. Addison-Wesley Professional. ISBN: 0321267974
[6] Jenney J. (2010) Modern Methods of Systems Engineering: With an Introduction to Pattern and
Model Based Methods. ISBN-13:978-1463777357
[7] OMG UML (2021) Unified Modeling Language version 2.5.1. Unified Modelling
https://www.omg.org/spec/UML/About-UML/ .
[8] Sajja, P.S., Akerkar, R. (2010) Knowledge-Based Systems for Development. Advanced
Knowledge Based Systems: Model, Applications & Research, Vol. 1
[9] UML Diagrams (2021) UML diagrams characteristic. www.uml-diagrams.org
[10] Veitaitė, I., Lopata, A. (2017) Transformation algorithms of knowledge based UML
dynamic models generation // Business information systems workshops BIS 2017, Poznan,
Poland, 28-30 June / editor Witold Abramowicz . - Series : Lecture notes in business
information processing. Vol 303. Cham : Springer International Publishing, 2017
[11] Veitaitė, I., Lopata, A. (2018) Problem domain knowledge driven generation of UML
models // Information and software technologies: 24th International Conference, ICIST 2018,
Vilnius, Lithuania, October 4–6, 2018: proceedings / edited by: Robertas Damaševičius, Giedrė
Vasiljevienė. Cham : Springer.
[12] Veitaitė, Ilona; Lopata, Audrius. Knowledge-based UML use case model
transformation algorithm // Business information systems workshops: BIS 2019 international
workshops, Seville, Spain, June 26–28, 2019: revised papers / W. Abramowicz, R. Corchuelo
(eds.). Cham : Springer, 2019. ISBN 9783030366902. eISBN 9783030366919. p. 39-48.
(Lecture notes in business information processing, ISSN 1865-1348, eISSN 1865-1356 ; Vol.
373). DOI: 10.1007/978-3-030-36691-9.
[13] Veitaite I., Lopata A. (2020) Knowledge-Based Transformation Algorithms of UML
Dynamic Models Generation from Enterprise Model. In: Dzemyda G., Bernatavičienė J.,
Kacprzyk J. (eds) Data Science: New Issues, Challenges and Applications. Studies in
Computational Intelligence, vol 869. Springer, Cham
[14] Veitaitė, Ilona; Lopata, Audrius. Knowledge-based UML activity model
transformation algorithm // Information society and university studies 2020 : proceedings of
the information society and university studies, 2020, Kaunas, Lithuania, April 23, 2020 / edited
by: Adrius Lopata, Vilma Sukackė, Tomas Krilavičius, Ilona Veitaitė, Marcin Woźniak.
Aachen : CEUR Workshop Proceedings. ISSN 1613-0073. 2020, p. 114-120. (CEUR
Workshop Proceedings, ISSN 1613-0073 ; vol. 2698).
[15] Veitaite I., Lopata A. (2020) Knowledge-Based Generation of the UML Dynamic
Models from the Enterprise Model Illustrated by the Ticket Buying Process Example. In:
Lopata A., Butkienė R., Gudonienė D., Sukackė V. (eds) Information and Software
Technologies. ICIST 2020. Communications in Computer and Information Science, vol 1283.
Springer, Cham. https://doi.org/10.1007/978-3-030-59506-7_3