Current
Students
Ph.D. Students
|
|
||||||||
M.Sc. Students
|
|
Graduated
Ph.D. Students
|
|
Caglar Durmaz, Ph.D., 2022 Position after
graduation: Founder and Manager at Integra Data
and Communication Technologies Co., Turkey Dissertation
Title: A Programming Language and Virtual
Machine for Developing Intermittently Powered Systems Dissertation
Abstract: In the last decade, the power
requirements of computing, sensing, and communication hardware have
decreased, and energy harvesting circuits have evolved. Microcontroller-based
devices have emerged that can operate without the need for batteries by
collecting energy from ambient sources such as solar, and radio frequency. A
battery-free device stores harvested energy in a small energy storage to
power the microcontroller and peripherals. When the stored energy is
exhausted, the device suddenly loses power. The device can only run again
after the storage energy reaches a threshold level. This phenomenon has led
to systems operating with intermittent power and the emergence of a new
computing paradigm called intermittent computing. Intermittent computing
requires special programming models to ensure that applications run correctly
despite power failures. However, current programming models lead to
hardware-dependent and non-reusable programs. This thesis is the first study
to introduce virtualization for the transiently powered computers to solve
these problems. A virtual machine called PureVM, which abstracts the
transiently powered computer, and a continuation passing programming language
PureLANG are introduced. With virtualization, intermittent power applications
are made portable and reusable. In evaluations between the PureVM virtual
machine and widely used programming model in the research area, PureVM
programs performed 25%-51% better in terms of runtime. |
|
|
Baris Tekin Tezel, Ph.D., 2020 Position after
graduation: Assistant Professor at Computer
Science Department, Dokuz Eylul University, Turkey Dissertation
Title: Debugging for the Domain-specific
Modelling Languages of Software Agents Dissertation
Abstract: Despite the existence of many agent
programming environments and platforms, developers may still face
difficulties in implementing Multi-Agent Systems (MAS). Therefore, both
working at a higher abstraction layer and modeling agent components can
facilitate MAS development. The most common way to achieve this level of
abstraction is to create Domain-Specific Modeling Languages (DSMLs) for MAS
development. However, although the integrated development environments of the
existing MAS DSMLs provide some kind of control over systems modeled
according to the syntax and semantic definitions of the respective language,
none of them have built-in support for debugging. This deficiency causes
agent developers to be unsure of the accuracy of the models prepared at the
design stage. To help filling this gap, this thesis presents a general
debugging framework that supports the design of agent components within
modeling environments. The debugging framework consists of 4 different metamodels
and a simulator. The use of the proposed framework begins with modeling a MAS
using a design language and converting the design model instances into a
runtime model. According to the framework, the runtime model is simulated in
a built-in simulator for debugging. The framework also provides a simulation
environment model and a control mechanism for simulation. The proposed
framework has been evaluated both qualitatively and quantitatively. All
results obtained were analyzed statistically. Evaluation results revealed
that, on average, the proposed framework reduces the time spent on debugging
by approximately 45% and increases the number of bugs repaired in models by
approximately 50%. |
|
|
Moharram Challenger, Ph.D., 2016 Position after
graduation: Postdoc researcher at Information
Technology Group, Wageningen University, the Netherlands R&D Director
of Unit Information Technologies Research, Development and Automation Systems
Co., Turkey Dissertation
Title: A Domain-specific Modeling Language
for Semantic Web enabled Multi-agent Systems Dissertation
Abstract: The need for realization of the many
different agent interactions and also agent properties such as autonomy,
responsiveness and proactiveness, make development of agent-based software
systems more complex than other software systems. Furthermore, the design and
implementation of internal agent behavior model and interaction within the
agent organizations of Multi-agent Systems (MASs) may become even more
complex and difficult to implement especially when the new requirements and
interactions for new agent environments such as the Semantic Web are taken
into account. In order to deal with the complexities of design and
implementation of such MASs, the use of a Domain-specific Modeling Language (DSML)
can provide the required abstraction which increases the descriptive power of
a MAS metamodel, defines the system semantics, and hence support a more
fruitful methodology for the development of MASs. Although agent researchers
have a great effort in MAS metamodeling and model-driven MAS development, a
significant deficiency exists in current studies when it is considered
providing a complete DSML for MASs especially working in Semantic Web
environment. In order to help
fill the abovementioned gap, a new DSML for MAS, called SEA_ML, is introduced
in this study which considers the architectural complexity of design and
implementation of MASs. To this end, a domain-specific metamodel is defined
that includes new viewpoints (in addition to the classical viewpoints of
MASs) which specifically support the development of software agents working
within the Semantic Web environment. This new metamodel paves the way for the
definition of an abstract syntax and textual/graphical concrete syntaxes for
the DSML of agent systems. Achieved DSML syntax is supported with both
textual and graphical modeling toolkits. The DSML also supports the
interpreter mechanism, so that, the framework can generate the architectural
code for different target platforms considering both the MAS and Semantic Web
aspects. In the meantime, both of the textual and graphical languages support
constraint control to validate the programs/models. In this way, some of the
semantic controls are realized in the design time of the program/model and avoids
potential errors which lead to more accurate code generation. Considering the
formal semantics of the DSML, the formal representation of SEA_ML’s semantics
are presented and used for MAS validation. The presented semantics defines
both static and dynamic aspects of the interaction between software agents
and semantic web services, in addition to the definition of the semantics
already required for agent internals and MAS communication. Finally, in this
thesis, an evaluation framework and systematic approach are also presented
for assessing existing or newly defined DSMLs for MASs. Evaluation of SEA_ML
according to this framework and systematic approach has been provided using
two evaluator groups. The evaluation is based on a multi case study research
approach and provides both qualitative evaluation and quantitative analysis.
The lessons learned are discussed considering the adoption of the evaluation
approach as well as the SEA_ML for supporting the generation of agent-based
systems. Evaluation results show that approximately 85% of the artifacts and
80% of the whole code is generated automatically by using SEA_ML. Also,
according to the evaluation results, the total time of development for both
groups shows that the group using SEA_ML needs half of the time the other
group needs for developing the case studies in average. |
Graduated
M.Sc. Students
|
|
Fahri Berkay Ercan, M.Sc., 2024 Position after
graduation: Embedded System Engineer at
CrossControl Co., Finland Thesis Title: A Low-code Development Platform for Embedded System Drivers and
Modules Thesis Abstract: The different software dynamics of kernel drivers and modules have
become an issue that developers have difficulty learning and spend time on.
These different software dynamics can be simplified through certain basic
skeleton structures and flow functions used on these structures. In this
thesis, it is aimed to create a more organized code base by creating the
mentioned basic structures and flow functions through a visual interface, to
enable the developer to get rid of the difficulties experienced in the initial
process of module development. At the same time, unlike general low-code
development platforms, the user's control over the code is left active
throughout the entire development process, removing the limitation on the
complexity of the created software. For quantitative evaluation of the
thesis, driver and module software available in Linux kernel driver version
5.15 were used. At the same time, a qualitative review and evaluation of the
platform proposed in the thesis was carried out according to the generally
accepted features in the relevant research field for low-code development
platforms. As a result of the evaluation, it was observed that the platform
complies with the basic low-code development platform features and that an
average of 30% of the existing code directories in Linux kernel drivers can
be created automatically. |
|
|
Cagatay Sahin, M.Sc., 2022 Position after
graduation: Software Development Senior Manager
at Veripark Yazilim Co., Turkey Thesis Title: A Comparative Evaluation of Using ASP.NET and Node.js Technologies
for Enterprise Web Application Development Thesis Abstract: In this study, a comparison of ASP.NET, which is one of the most
preferred technologies for the development of corporate-scale web-based
applications, and Node.js technologies, the rate of which is increasing the
fastest, was performed. Firstly, a set of criteria was created for the
comparison, and sample applications were developed using both technologies.
As a result of the experiences gained from these studies, both a qualitative
comparison according to the determined criteria and a quantitative analysis
were carried out in line with the measurements. According to the results
obtained, it can be said that both technology platforms have the competencies
to meet all the requirements to develop a business application within an
enterprise scale. It has been examined that the Node.js platform is superior
in criteria such as easy accessibility of application development tools and
diversity of libraries. The superiority of ASP.NET platform has been observed
in criteria such as ease of learning, rapid program development and
robustness. Run time performances of the sample programs were also compared,
and it was observed that the Node.js applications were running faster, although
the resource consumption of both platforms were measured at a similar level.
It is believed that the comparison given in this thesis may give software
developers an idea about which technology would be a more appropriate choice
under which criteria and under which conditions. |
|
|
Alper Silistre, M.Sc., 2021 Position after
graduation: Software Engineer at Portera Co.,
the Netherlands Thesis Title: An End-to-End Scalable Approach for Model-based Graphical User
Interface Testing Thesis Abstract: Model-based GUI testing has been a subject of many studies for
decades. It has also started to gain attraction in the commercial field in
the last years. It is part of black box testing due to its ability to test
systems without accessing the codebase. Since model-based GUI testing needs
only a model for testing, it can be automated easily in comparison to the
code-based testing approaches. This thesis study addresses the scalability
problem in systems with big models when model-based GUI testing is expected
to be used. Sometimes it takes so much time to generate and execute test
cases due to the model's complexity. An approach is presented and exemplified
with two case studies. A community detection algorithm is applied, and
communities are detected to split the model into smaller models. Test cases
generated from the primary and community models are compared. In order to
validate the approach's effectiveness, mutation testing was applied to
calculate mutation scores. Test suite sizes of models reduced dramatically
with a community detection approach without compromising the mutation score
too much. An open-source software is developed to apply the approach in an
automated way. The use of this software kept the manual effort as low as
possible. Results and possible improvements are discussed to increase the
capabilities of the model-based testing tool. |
|
|
Tansu Zafer Asici, M.Sc., 2020 Position after
graduation: Software Developer at Unipa Co.,
Turkey Thesis Title: On the Use of the Analytic Hierarchy Process in the Evaluation of
Domain-specific Modeling Languages for Multi-agent Systems Thesis Abstract: Software agents and multi-agent systems (MAS) composed by these agents
are used in the development of the complex intelligent systems. In order to
facilitate MAS software development, various domain-specific modeling
languages (DSMLs) exist. Unfortunately, the usability evaluation of these
languages are mostly not considered or only a few assessments which cover one
single MAS DSML are made. In this thesis, a comparative MAS DSML evaluation
methodology based on the Analytical Hierarchy Process (AHP) is introduced. A
categorized set of MAS DSML criteria which can be used for the multi-criteria
decision making is defined. These criteria can be prioritized by the
developers according to their modeling language expectations and the
application of the methodology allows the evaluation of DSML alternatives
based on this prioritization. As the result, the most appropriate DSML is
determined. The proposed methodology was applied for the comparative
evaluation of four well-known MAS DSMLs. The conducted evaluation showed that
the agent developers prioritized appropriateness, completeness and shortening
the development time as the most significant criteria while the attractiveness
of the notations had a minimum effect on preferring a language. Favourite
DSML for each comparison category and criteria was determined. |
|
|
Hussein Masoud Marah, M.Sc., 2020 Position after
graduation: Ph.D. Student at Johannes Kepler
Universitat Linz, Austria Thesis Title: A Model-driven Round-trip Engineering Methodology for TinyOS
Applications Thesis Abstract: In this thesis, a round-trip engineering (RTE) methodology is introduced
both for the model-driven development (MDD) of the Wireless Sensor Network
(WSN) applications executed on the well-known TinyOS operating system and the
regain of system models from the existing TinyOS code. The first part of the
methodology enables the model-driven engineering (MDE) of TinyOS applications
by using a domain-specific modeling language (DSML), called DSML4TinyOS. In
the second part of methodology, a reverse engineering mechanism is applied on
the codes of an existing TinyOS application to regain the corresponding
software model without any human intervention. These models can also be
processed in the graphical modeling environment provided by the DSML4TinyOS
language. An evaluation of the proposed methodology and its tools was performed
by taking into account WSN applications which are publicly available in the
TinyOS GitHub repository. Approximately 40% of TinyOS Module components and
100% of the TinyOS Configurations were automatically generated by only
modeling with DSML4TinyOS language. Moreover, complete code of these
applications existing in the GitHub distribution was given into the RE4TinyOS
platform and the corresponding graphical models of these applications with
including all major components of the programs were successfully regained and
processed inside the DSML4TinyOS environment. |
|
|
Mustafa Abshir Mohamed, M.Sc., 2020 Position after
graduation: Independent Researcher and Software
Developer at SomDigital Co., Somalia Thesis Title: Model-driven Engineering of Cyber-Physical Systems Software: A
Systematic Literature Review Thesis Abstract: Development of Cyber-Physical Systems (CPS) draws more interest from
both researchers and practitioners considering the opportunities they offer
in almost all areas of industry. However, development and management of CPS
are challenging tasks due to their inherent heterogeneity and complexity
characteristics. Regarding the development of CPS, there currently exists no
standard methodology owing to the complexity of the domain. One of the key
approaches to reduce the development complexity for CPS is Model-driven
Engineering (MDE), which is frequently used in many business domains for
software development to increase the level of platform abstraction. This
thesis presents a Systematic Literature Review (SLR) study that focuses on
identifying and classifying the recent research practices pertaining to CPS
development by applying MDE approaches. The study evaluates 140 research
papers published during years 2010–2018. Accordingly, a comprehensive
analysis of various MDE approaches and tools used in the development
life-cycle of CPS is provided. Furthermore, the study identifies the research
gaps and areas which need more investigation. Conducted study may help
researchers and practitioners to get an overall understanding of the research
trends and existing challenges for further research and development. |
|
|
Ufuk Firtina, M.Sc., 2020 Position after
graduation: Business Analyst at Vakifbank Asset
Management Co., Turkey Thesis Title: Extending DSML4BDI Language with JaCaMo Platform Thesis Abstract: In Agent-oriented Software Engineering (AOSE) research field, various
domain-specific modeling languages (DSMLs) are introduced for the development
of Multi-agent Systems (MASs). It is important to design the internal
structures of the agents according to architectures such as
Belief-Desire-Intention (BDI), to create execution environments that can be
used by the agents in the MAS system and to define the organizational
structures of the agents. Within this context, DSML4BDI is one of the DSMLs
proposed for AOSE. The language, which has a metamodel conforming to the
Jason platform, supports the modeling of the internal structures of the
agents. However, provided modeling elements are insufficient especially for
the creation of MAS, i.e. developers are struggling both to create common
execution environments which agents will use and to define organizational
structures. In this thesis, a DSML, called DSML4BDI 2.0, is developed which
eliminates the aforementioned deficiencies of DSML4BDI and increases the
execution platform richness of the created models. A metamodel is designed to
include Agent, Environment and Organization viewpoints for DSML4BDI 2.0. A graphical
concrete syntax is derived to create MAS models conforming to this metamodel.
Hence, developers can model the MASs to be implemented by using this
graphical notation. Based on the semantics provided from model to text
transformations, DSML4BDI 2.0 models can be run on the JaCaMo platform which
supports environment programming. Thus, it is possible to generate code
output from the models created with DSML4BDI 2.0. Finally, an evaluation of
DSML4BDI 2.0 is carried out by analyzing the results obtained from various
MAS development case studies. |
|
|
Egemen Evin, M.Sc., 2019 Position after
graduation: Software Developer at MEX
International Group Co., Turkey Thesis Title: Design and Implementation of a Software for the Assessment of
Eco-Driving Performances of Public Transport Drivers Thesis Abstract: Driving vehicles according to eco-driving principles and techniques
have significant impact on decreasing both fuel consumption and CO2
emissions. In addition to some kind of technical and/or mechanical features
brought by today’s new generation vehicles, driver behavior is also one of
the greatest factors affecting the fuel consumption. Many studies show that
the effect of eco-driving education on the drivers loses its impact in long
term and there needs some sort of continuous monitoring and/or feedback
mechanisms. Moreover, in order to adapt behavior to drive more economically,
information about instant fuel and CO2 waste has to be provided to
the driver. Taking into account above mentioned requirements, in this thesis,
an eco-driving system in which data gathered from the CANBus of public
transport vehicles are processed for both comparative and fair evaluation of
bus drivers’ eco-driving performance. Moreover, in-vehicle components of the
system guide the drivers during their trips and provide feedbacks and
real-time warnings considering their current eco-driving performance.
Developed system was successfully deployed and evaluated in a public
transportation system used by approximately 250 thousand passengers daily.
The results are very promising in the sense that both drivers and operators
found the system useful and the system provided fuel saving. |
|
|
Omer Faruk Alaca, M.Sc., 2019 Position after
graduation: Software Engineer at Kron
Information Technology and Services Co., Turkey Thesis Title: A Tool Supporting the Systematic Evaluation of Domain-Specific
Modeling Languages for Multi-Agent Systems Thesis Abstract: In Agent-oriented software engineering (AOSE) research field, various domain-specific
modeling languages (DSMLs) are proposed for the development of Multi-agent
Systems (MASs). These DSMLs are supposed to meet software development
requirements of both MAS DSML developers and MAS DSML users. Moreover, they
should be evaluated to determine how they both speed up and facilitate MAS
development. However, many MAS DSML studies do not consider such an
evaluation of language components and generated artifacts over using the
language while design and implementation of agent software. In order to fill
this gap in AOSE research, a web-based tool is developed in this thesis. The
tool provides the systematic evaluation of MAS DSMLs from both quantitative
and qualitative aspects according to a series of pre-defined metrics. During
the quantitative evaluation inside the tool, MAS domain coverage is
determined by comparing a DSML’s metamodel with a reference MAS metamodel. In
the second part of the quantitative evaluation, which is constructed on a
multi-case study, the software development time and the artifact generation
performance are taken into account. Finally, qualitative evaluation of the
language is performed in which MAS developers are requested to answer a
questionnaire and assess the language according to some quality metrics. Use of
the proposed online MAS DSML evaluation process is exemplified in the thesis
with the evaluation of Prometheus/PDT, a well-known tool in AOSE. |
|
|
Huseyin Kucuk, M.Sc., 2019 Position after
graduation: Software Developer at Masomo Gaming
Co., Turkey Thesis Title: A Reverse Engineering Methodology for the Development of
Belief-Desire-Intention Agents Thesis Abstract: Model-driven architectures, used for the development of
Belief-Desire-Intention (BDI) agents, are mostly constructed by following a top-down
approach in which software agents are modelled according to a metamodel and
the prepared models are given into a series of model-to-model and
model-to-text trasnformations to automatically generate the agent
implementations. However, modifications made in these auto-generated
artifacts can not be reflected back to the agent models. Hence the
synchronization between the agent models and the corresponding software is
ruined which also makes the reusability of agent models and its software
difficult. In order to eliminate these deficiencies, a reverse engineering
methodology and a supporting tool for the automatic generation of BDI models
from existing agent software are introduced in this thesis. With the use of
the proposed tool, it is possible to both re-generate BDI agent models from
the programs running on the well-known JACK platform and reflect the changes
made in the programs to the corresponding models. |
|
|
Berkay Akdal, M.Sc., 2018 Position after
graduation: Software Developer at Galaksiya Information
Technologies Co., Turkey Thesis Title: A Domain-specific Language for the Model-driven Analytics of
Logistics Data Thesis Abstract: Elasticsearch is a distributed RESTful search engine, capable of
solving growing number of use cases and can handle petabytes of data in
seconds. However, Elasticsearch comes with a complex query language which
causes a steep learning curve for the developers and, therefore, creation of
queries can be difficult and time-consuming in many cases. Hence, in this
study, a domain-specific modeling language, called Dimension Query Language
(DQL) has been developed, to support the model-driven development of
Elasticsearch queries. Elasticsearch queries can be automatically generated
from DQL models and DQL’s IDE is capable of executing these auto-generated
Elasticsearch queries on remote repositories. Specifically, the logistics has
been determined as the target business domain and capabilities of using DQL
have been demonstrated for the analytics of logistics data which are very
large in volume. An evaluation of using DQL has been performed at the
industrial level with the participation of a group of developers. The
conducted evaluation showed that the use of the language significantly
decreases the development time required for creating Elasticsearch queries.
Finally, qualitative assessment, based on the developers’ feedback, exposed
how DQL facilitates the development of Elasticsearch queries. |
|
|
Sadik Arslan, M.Sc., 2018 Position after
graduation: Electronic Engineer at Kentkart Ege
Electronics Automatic Fare Collection & Vehicle Tracking Systems Co.,
Turkey Thesis Title: Model-driven Development of Device Tree Software Thesis Abstract: Device Tree (DT) files, which include hardware information and
configuration, are used during the compilation of operating systems for
various embedded platforms. DTs provide description of physical devices and
peripherals inside an embedded system hardware with node specifications.
However, software developers mostly have difficulties in developing such
applications due to text-based syntax of DT source files which has a
different structure from the well-known general purpose programming
languages. Furthermore, the developer needs to be familiar with the hardware
which is specific for each different microprocessor to prepare such DT files.
In order to eliminate these deficiencies and difficulties of current DT
software development processes, a model-driven software development
methodology in which automatic generation and configuration of DT software
for different embedded system platforms is provided in this thesis. Also, a
domain-specific modeling language, called DSML4DT, is developed with
including all supporting visual modeling and automatic code generation tools for
the application of the proposed methodology. An embedded system device is
used for evaluation. The comparative evaluation results showed that DT
software can automatically provide a significant amount of the required code
using DSML4DT. |
|
|
Emine Bircan, M.Sc., 2017 Position after
graduation: Computer Engineer at TUBITAK
National Research Institute of Electronics and Cryptology (UEKAE), Turkey Thesis Title: Horizontal Transformations between Platform Independent Metamodels of
Software Agents Thesis Abstract: The conventional approach currently followed in the development of
domain-specific modeling languages (DSMLs) for multi-agent systems (MASs)
requires the definition and implementation of new model-to-model and
model-to-text transformations from scratch in order to make the DSMLs
functional for each different agent execution platform. In this thesis, an
alternative approach, which considers the construction of the
interoperability between MAS DSMLs for a more efficient way of platform
support extension, is proposed. The feasibility of using this new
interoperability approach instead of the conventional approach is exhibited
by discussing and evaluating the model-driven engineering required for the
application of both approaches. Use of the approaches are also exemplified
with a case study which covers the model driven development of an agent-based
stock exchange system. In comparison to the conventional approach, evaluation
results show that the interoperability approach requires both less development
time and effort considering design and implementation of all required
transformations. |
|
|
Miray Tosun, M.Sc., 2015 Position after
graduation: Software Engineer at R&D Center
of OBASE Computer and Consultancy Co., Turkey Thesis Title: A Methodology for the Automatic Generation of JavaCard Software Thesis Abstract: Java Card platform enables the object-oriented design and
implementation of smart card applications. However, only a subset of Java
programming language can be used due to limited processing and storage
capabilities of these devices. Further, developers are forced to deal with
the constraints and details of a low-level communication protocol for the
packet transmission between smart cards and host computers. In order to eliminate
those difficulties of Java Card software development and support the
model-driven development of on-card applications, a domain-specific language,
called DSL4JavaCard has been developed in this thesis study. The definition
of both a metamodel for the language's abstract syntax and a concrete syntax
for checking some contraints during smart card software modelling in the
thesis paves the way of automatic generation of Java Card application codes
from DSL4JavaCard instance models. In addition, case study-based
exemplification of the language usage and evaluation are also provided in the
thesis. Achieved results show that DSL4JavaCard differentiates from the
previous related work specifically by enabling the model-driven development
of Java Card security components. Moreover, it is experienced that
operational semantics of the language succeeds in automatic generation of all
fundamental Java Card components required in a smart card application while
providing the automatic generation of most of the user-defined components and
code blocks specific to the bussiness domain of the smart card application. |
|
|
Kemal Deniz Teket, M.Sc., 2013 Position after
graduation: Ph.D. Student at Laboratoire d'InfoRmatique
en Image et Systèmes d'information (LIRIS), Institut national des sciences
Appliquées de Lyon (INSA Lyon), France Thesis Title: Agent-based Parent Selection in Peer-to-Peer Mesh-based Video
Streaming Systems Thesis Abstract: Research on the use of software agents in Peer-to-Peer (P2P) video
streaming systems has recently emerged in order to meet the requirements such
as time constraints, minimizing delay and efficient utilization of network
resources. However parent selection and network awareness are not addressed
properly in these leading studies. In order to fill this gap, design and
implementation of a Multi-agent System (MAS) for P2P video streaming are
introduced in this thesis. The construction of the MAS by applying a
well-defined agent-oriented software engineering methodology is discussed.
The internal structure of each agent in the proposed MAS is based on the
well-known Belief-Desire-Intention (BDI) architecture. Experimental results
show that the proposed MAS succeeds in the related parent selection and
promotes the system execution. BDI-based MAS design given in this work can
guide those researchers who aim to implement autonomous mesh-based P2P video
streaming systems. |
|
|
Orcun Yildirim, M.Sc., 2012 Position after
graduation: Industrial Engineer at Cimentas
Group, Turkey Thesis Title: A Mobile Agent-based Decision Support System for Cement Industry Thesis Abstract: Nowadays, the World is globalized and borders are nearly removed between
countries. So, the competition between companies is increasing. As a
solution, companies are trying to decrease their costs. Production planning
is very important for decreasing the costs and like other industries, it has
also become critical in cement industry within this direction. In this
thesis, a cement production planning software system which tries to minimize
energy cost and meet some constraints is presented. The mobile technology and
software agents are used while designing the system. The production plan is
prepared by using messaging properties of the agents as well as their
autonomous structure for communication between departments. The plan reduces
energy cost and covers all production constraints. As a result of the new agent system, a serious drop at the
energy cost was observed. At the same time, automation reduced the workload
of employees significantly. So, companies will have an advantage at this
competition. |
|
|
Sinem Getir, M.Sc., 2012 Position after
graduation: Ph.D. Student at Institute of
Software Technology, Universität Stuttgart, Germany Thesis Title: Semantics of a Domain-specific Modeling Language for Semantic Web
enabled Multi-agent Systems Thesis Abstract: Multi-agents Systems (MAS) are widely used in many applications of
various areas such as robotics, air traffic, space, security applications and
energy. Therefore there exist various agent frameworks or middlewares for
MASs (e.g. JADE, JACK, JADEX and SEAGENT) which are mostly based on
General-purpose Programming Languages such as Java. However a user who
defines the requirement analysis does not usually need to know specific
programming structures of complex MASs. On the other hand, internal structure
of agents and interactions belong agents may become even more complex to
implement when new requirements for new agent environments such as the
Semantic Web which improves the current world wide web (WWW) are taken into
account. SEA_ML is a domain specific
language that aims to provide the required abstraction and supports a more
fruitful methodology for MASs on Semantic Web independently from the agent
platforms which are listed above and in the literature. In this thesis, both
denotational and operational semantics are defined for SEA_ML. Alloy language
which is a declarative language based on first order and relational logic is
used to define the formal semantics of SEA_ML. SEA_ML model analysis is
completed with Alloy tool and implementation of semantics is provided with
model transformations by using ATL. |
|
|
Suleyman Yildirim, M.Sc., 2012 Position after
graduation: Post-M.Sc. Software Designer Trainee
at 3TU.School for Technological Design, Stan Ackermans Institute, Eindhoven
University of Technology, the Netherlands Thesis Title: An Agent Architecture for Peer-to-Peer Live Video Streaming
Applications Thesis Abstract: Software agents bring a promising approach for solving the complexity
and they have been widely used in many application domains. As the agents
exhibit autonomous, reactive and proactive behavior, they are also useful for
solving various problems of peer-to-peer video streaming. In this study, a
novel agent architecture based on the well-known Belief-Desire-Intention
(BDI) model is designed in order to increase the video quality of the peers
in video streaming applications. The proposed architecture is applied in
partner selection problem which is an important area in video streaming. An
agent-based software system is developed. The video bitrate, delay and hop
count from source to destination nodes are given to the software agents as
environment facts and agents make partner selection according to these facts.
Two different agent behaviors are designed in the proposed architecture,
namely reactive and proactive behavior. It is aimed that the proactive
behavior makes better decisions when compared to reactive behavior. The test
results show that both proactive and reactive behaviors of the proposed
architecture manage to increase the video quality by keeping the total received
bitrate value above the threshold. Moreover, it is observed that the
proactive behavior applies better decisions by exhibiting goal-oriented
behavior when compared to reactive behavior. |
|
|
Sebla Demirkol, M.Sc., 2012 Position after
graduation: Software Developer at Veripark Co.,
Turkey Thesis Title: Syntax of a Domain-specific Modeling Language for Semantic Web
enabled Multi-agent Systems Thesis Abstract: Agents and Multi-agent systems (MASs), which consist of the agents,
are used as substantial technologies in order to develop complex systems with
autnomous and reactive behaviour. In this sense, various software frameworks
exist intended for improving agent oriented software development. However,
semantic web is not supported by most of these software frameworks. The
semantic web provides the interpretation of web page contents through
ontologies by machines and represents the new generation of World Wide Web.
The current software development frameworks which support the collaboration
of agents with semantic web elements are complicated enough in order to be
used by system developers who are not expert on the related domain.
Furthermore, users, who define system requirements, usually do not need to be
aware of the structures specific for agent programming. In this thesis,
graphical and textual syntaxes are defined for a domain specific modeling
language (DSML) for MASs which specifically supports the interaction between
agents and semantic web services on the Semantic Web environment. In order to
exemplify the use of the related DSML, code generation from defined syntax is
discussed. |
|
|
Hidayet Burak Saritas, M.Sc., 2011 Position after
graduation: Software Developer at Kentkart Ege
Electronics Automatic Fare Collection & Vehicle Tracking Systems Co.,
Turkey Thesis Title: Model-based Engineering for the Development of Smart Card Software Thesis Abstract: Smart cards are portable integrated devices that store and process
data. Speed, security and portability properties enable smart cards to have a
widespread usage in various fields including telecommunication,
transportation and credit card industry. However, the development of smart
card applications is a difficult task due to the hardware and software
constraints. The necessity of the knowledge of both a very low-level
communication protocol and a specific hardware causes smart card software
development to be a big challenge for the developers. Written codes tend to
be error-prone and hard to debug because of the limited memory resources.
Hence, in this study, a model-based approach which facilitates the smart card
software development is introduced. Proposed approach provides easy design of
smart card systems and automatic generation of the smart card software from
the system models. A Platform Independent Metamodel (PIMM) is created based
on the smart card standards in order to support easy development of instance
models and automatic code generation with model transformations for two major
smart card platforms: Java Card and Basic Card. |
|
|
Ferhat Erata, M.Sc., 2010 Position after
graduation: Co-founder and president of Unit
Information Technologies Research, Development and Automation Systems Co.,
Turkey Thesis Title: A Platform-specific Metamodel for Developing Web-based Systems Thesis Abstract: Various Web engineering methodologies and metamodels exist to design
and develop web applications for systematic design followed by an automatic
generation of Web applications. Without a specific platform technology,
automatic generation is not efficient and productive in terms of meeting
business requirements. On the other hand, such technologies are not enough
abstract to make transformations and mappings from them straightforward. In
this thesis, a platform specific web engineering metamodel is proposed. A
method based on Model Driven Architecture and a framework built on top of
Microsoft SharePoint application framework are both proposed to perform
generation of web based enterprise systems from user models of the metamodel.
The proposed metamodel divides content-driven separate concepts of Windows
SharePoint Services Object Model (WSSOM) into viewpoints and increases
abstraction level by means of redefinition of the concept relationships. |
Graduated
M.Sc. Non-Thesis Students:
|
|
|
|
Selime Saglam, M.Sc., 2024 Position after
graduation: Software Developer at TUBITAK BILGEM,
Turkey Graduation
Project Title: An Investigation on the Model
Execution Feature of Eclipse Papyrus Modeling Environment |
|
|
Petek Savas Hamitbeyli, M.Sc., 2022 Position after
graduation: Frontend Developer at Modaramo Media
Co., USA Graduation
Project Title: Development of an E-Commerce
Application using the MERN Stack |
|
|
Ferec Hamitbeyli, M.Sc., 2022 Position after
graduation: Mobile Application Developer at PATH
Product and Software Development House Co., Turkey Graduation
Project Title: Design and Implementation of a
Running Analysis Application on Android Operating System using Kotlin
Programming Language |
|
|
Onur Yildirim, M.Sc., 2022 Position after
graduation: Game Developer at Tiplay Studio Co.,
Turkey Graduation
Project Title: Design and Implementation of an
iOS-based Mobile Conference Management System using Swift Programming
Language |
|
|
Gokmen Balkan, M.Sc., 2018 Position after
graduation: Software Developer at Univera Co.
R&D Center, Turkey Graduation
Project Title: Applying Design Patterns to
Procedural Generation of Levels in Role Playing Games |
|
|
Hamit Unugur, M.Sc., 2014 Position after
graduation: Software Developer at MAN Turkey
Co., Turkey Graduation
Project Title: Assessment of the Semantic Service
Discovery Methods on Mobile Platforms |
|
|
Ozge Ozkaya, M.Sc., 2013 Position after
graduation: Responsible of Ege University
International Computer Institute’s PC Laboratory, Turkey Graduation
Project Title: A Study on the RESTFul Web Services |
|
|
Dincer Can, M.Sc., 2013 Position after
graduation: Information Technology Teacher at
Republic of Turkey - Ministry of National Education, Turkey Graduation
Project Title: A Software that Records and Displays
Optical Disc Contents |
