Regression models are undoubtedly one of the most frequently used techniques in applied statistics. Prognostic regression models have been a contributing factor in the evolution of various research fields, one of which is genetics. The use of the new genotypic technologies helped to increase the amount of genetic data, which are used in Genome Wide Association Study analysis to explain the phenotypic variation. Due to the vast amount of data, however, it is impossible to study the interactions between the genetic loci. The purpose of this study is the development of an algorithm that will essentially, though various criteria, reduce the number of pairs of genetic loci to be tested for interaction and the best possible usage of the computing system resources, in order to achieve a realistic completion time of the analysis.

Deep machine learning models are used in various applications such as spam email detection and face recognition for user authentication. However, these applications also provide incentives for adversaries to perform attacks where they interfere with the data related to the training, deployment and running of the models for malicious purposes. Several of such actions fall under the category of Data Poisoning attacks. These attacks can lead to biased and incorrect predictions by the machine learning model. Currently, there is no universal solution that can effectively defend against all types of Data Poisoning attacks. The present doctoral research aims to investigate vulnerabilities of machine learning models when it comes to Data Poisoning attacks, as well as their potential defenses and mitigation methods. Additionally, it will examine and compare solutions that have been proposed in the past. The expected contribution of this research is the development of a solution that can enhance the defense of machine learning models against Data Poisoning attacks, with a focus on higher effectiveness, scalability, and robustness compared to previous solutions.

Each science conducts its own research in accordance with its own rules, principles, and priorities. An object can be the subject of study for different disciplines, providing evidences, data, and information, useful for everyone. The Antikythera Mechanism is a characteristic example of an artifact that has been the “holy grail” of research and study by various sciences, such as Engineering, Astrophysics, Archaeology, History, and Informatics. Each one examines the Antikythera Mechanism from its own perspective and accumulates new information in existing knowledge. For many, the Antikythera Mechanism is the first analog computer to be saved in its original form. The synthesis of all the data and information gathered from different disciplines to illustrate the complex functions that the Antikythera Mechanism performed, using new technologies and advanced software tools, is the challenge of the doctoral thesis. The results of the present research will be used as a feedback to the scientific community to continue research on the subject, while enabling the simple user, through high-level immersive experiences in virtual reality, in the context of serious 3D games, for better understanding and communication of complex and emblematic artifacts of the World Cultural Heritage.

This dissertation examines the broader context of Information Science from the perspective of Smart Education. The evolution of technologies as well as the daily requirements, lead the wider field of education in the use of applications for learning and research. It is important to consider this new approach to learning and whether it affects educators and apprentices in their performance. The aim of the dissertation is to highlight technologies that are constantly evolving and adapting to new data that are rapidly updated in recent years, that contribute to the development of Smart Education. The Internet of Things combaning with various technologies, improves the storage, management, analysis and transfer of Big Data that require the existence of Cloud Computing. An important issue worldwide is the protection of the personal privacy of users, who generate data and share them for various purposes. So, Internet of Things, Big Data and Cloud Computing technologies are investigated focusing on security issues that may arise. The expectation of the above combination is to bring accurate analysis and the best possible results, for a modern and more sustainable education environment.

A. Introduction

 This PhD dissertation proposal aims at studying, designing, developing and evaluating an integrated work framework for best activity based costing in business process.

More emphasis will be given to the use of engineering techniques of learning and optimization techniques to identify cost centers for the different types of sources of entrepreneurship.

 The main types of sources used in business as cost centers are:

(1) Natural resources (production units, distribution centers, warehouses, etc.), (2) human resources (different types of staff, etc.), (3) the financial position (cash flows, limit of borrowing, unsecured accounts, etc.), (4) technology issues (ERP, etc), (5) promotion issues (promotional expenditure, promotional campaigns, etc.), (6) management issues (corporate culture, education system, etc.) and (7) legal issues (patents, rights of use, signing of contracts, etc.).

 B. Objectives – Contribution

 The proposed research aims to study fields of research that seek a more efficient use of activity based costing. In particular, emphasis will be placed on identifying cost centers and distributing the cost of a service or product using engineering and optimization techniques. In particular, emphasis will be placed on identifying cost centers and distributing the cost of a service or product using engineering and optimization techniques. Until now, such techniques have been used to select a particular cost center rather than many.

It will also evaluate the proposed methodology in a case study involving a real enterprise.

The proposed research will contribute to its more efficient use of activity based costing by using engineering and optimization techniques.

The main goals of this research are:

-         To study the Gamification Technologies and Gamification Underlying Theories and game-based methods and strategies for the design of digital training environments in business and organizations

-          To develop a methodology that can help us to implement  tools, and methods around designing gamification in e-commerce environment and digital training environments. For the purposes of this research, the implementation of case study projects includes the design of gamified activities for educational seminars. The digital material of the first educational seminar sets client service into the spotlight with the proper configuration of the digital environment of the Moodle platform and is addressed to adult trainees. The digital material of the second educational seminar is designed to support development of entrepreneurial mindset in university students with game-based methods and strategies and the proper configuration of the digital environment of the WordPress platform. 

-          To design and develop a general framework for personalized gamification and centre this framework on the core elements of a recommender system.

-     To implement a case study project of the above-mentioned methodology and general framework with main goal the personalized gamification

The Ethereum blockchain network was created with the central idea of not just being a medium of exchange in unreliable environments and without intermediaries, but of developing an ecosystem in which technologies such as Decentralized Applications (DApps) took shape and functionality. A key component of the structure of these apps are smart contracts, code programs stored within the Ethereum network that run automatically when certain conditions are met. Inside their code we are given the opportunity to integrate the business logic of DApps, thus allowing the realization of reliable data exchanges, agreements and monetary value.

But in many cases creating a proper and secure contract ends up being a difficult task for both novices and experienced developers. Poor code reuse, different programming styles and incomplete or even incorrect application of software design patterns could lead to huge financial losses. To address the above problems, an online code recommendation framework is proposed as an artifact of the Design Science Research methodology, in which an organized set of established software design patterns will facilitate the work of developers to select and implement all the necessary design patterns during the creation of a smart contract using correct, functional and secure programming solutions.

Intelligent Transport Systems (ITS) is a modern field of research dealing with the development and provision of innovative services related to transport and traffic management systems. These systems involve both terrestrial and aerial means and their study aims at addressing issues of safety, efficiency and sustainability of transport networks, reducing congestion and improving the driving experience. More specifically, ITS is implemented by the collaboration of advanced technologies, including electronic sensors, data transmission systems, and intelligent control implementation mechanisms. The aim of these systems is to provide advanced services and improve the experience of drivers and passengers, such as, for example, the provision of information on public transport or information on congestion.

Big data analysis has played a key role in the development of ITS. The methods used in big data analysis are capable of solving problems of data storage, management and processing. It therefore becomes an important tool for improving the efficiency of its operation and increasing the level of safety. In particular, the characteristics of big data analytics in ITS include advanced transport management, advanced traveler control and information systems, commercial vehicle management and advanced public and urban transport systems.

Medical imaging processing is a multidisciplinary research field used for disease`s early diagnosis and efficient treatments search, among many other fields. Common types of medical imaging are Computed Tomography (CT), Magnetic Resonance Imaging (MRI) scanner or can be X-rays, Nuclear medicine imaging including positron-emission tomography (PET) and ultrasound. In most processing applications difficulties arise due to the existence of noise, which can lead to false conclusions with affecting models robustness, and diagnostic processes. There are several noise sources and types which depend on the type of medical imaging, including but not limited to, radiofrequency emissions, quantum noise, and occlusions. In this thesis, we investigate the applicability of synergetic schemes towards data harmonization and outliers’ removal using AI and deep learning tools, aiming at improving the diagnostic capacity of deep learning models.

The modern way of life requires the exchange of large amounts of information via
telecommunication networks. The demand for high transfer rates is steadily
increasing, while the recent pandemic is weighing on the situation. Optical backbone
networks serve much of the information being transferred. Although a reliable
transmission medium, fiber optics are at risk of failures due to natural and
technological disasters, with large impact on the function of the state, businesses, and
the needs of citizens. Passive optical networks are a fast-growing access technology
for high-speed connection of end-users to backbone networks, having constant
research interest. The purpose of the doctoral research is to study functional
requirements of passive optical networks, namely availability, survivability, and energy
efficiency, and to produce a methodology based on formal analysis and verification.
The research will record possible failures of optical networks and examine disasters
that cause them. The formal analysis and verification of the failure consequences will
be based on an initial model of functional requirements which will be refined at a
second stage. Metrics for assessing the operation of passive optical networks will be
evaluated. The results will be used to implement simulations and exploited towards
improving the operational requirements defined by this research.

The rapid development of e-commerce worldwide, and in particular digital platforms, is the basis for the preparation of this doctoral thesis. Its purpose is initially to identify the different types of digital platforms that are emerging in the global market and to analyze their external and internal environment. Then the factors that influence their effectiveness will be searched and various methods of measuring them will be presented. This work will be done in the light of internationalized markets and a global environment.

Business continuity processes is vital for any organization, because stopping them automatically means huge economic, technological and social disaster.  There are many examples of such a disasters that have cost to organizations their survival and viability, such as the most recent Covid-19 pandemic.  Business Continuity used to be treated by organizations as a commitment to invest resources that cannot be used and their value or return cannot be assessed, so they considered it unnecessary, since the  probability of occurrence was extremely low [V.  Cerullo, M. Cerullo 2006].  Over the years and when it was found that ad hoc disaster or emergency management was inefficient and created huge problems due to the lack of a specific plan, which had multifactorial extensions.  Following this finding, they continued to adopt business continuity procedures (BCPs).  Nowadays a percentage of organizations have complied and even commit enough funds for its proper operation and efficiency.  But as there is a tremendous technological advancement and organizations are following the digital transformation, the BCP should be studied with the adoption of emerging technologies, globally.

The modern theories of Business Administration consider business processes as the core asset of an enterprise and claim that their organizational structure must be built upon processes rather than administrative hierarchies. This approach emerges the development of Business Process Management, which combines research from business administration and computer science in order to invent methods for the analysis, design and optimization of processes. The latter is an NP-hard problem due to the non-linear, non-convex and discontinuous nature of the involved mathematical models and is defined as the problem of constructing feasible process designs under specified requirements and their comparison in the base predefined evaluation criteria. Two promising approaches for the confrontation of the process optimization problem are the Evolutionary Computing (EC) due to the adaptability, the robust performance and its global search character and the Constraint Logic Programming (CLP) which is by nature a searching technique in the domain of decision variables under constraints. The incorporation of these methodologies in a single system could combine the generation of several parts of the solution in a single run offered by EC, with the embedded backtracking mechanism of CLP that permits the selective restart of an algorithm from stored branching points

The use of Chatbots in education is at a fairly early stage worldwide, and it seems that it can be an important tool for teachers and students over the next few years. The above field is presented in the literature, as a tool which combines the traditional learning with the modern one (blended learning). In this way, it maintains the interactivity, which is lacking in asynchronous distance learning, supports students without a time limit, provokes a sense of pleasure in its use, while reinforcing students' motives for learning and giving them the opportunity to self-disclosure.The implementation of such a modern tool in the educational processis a challenge which requires further investigation. Before being used in the learning process, it should be considered for acceptance by teachers, students and the educational system. The benefits of its use should also be investigated, as the possibility of integrating it into the modern classroom.

Metaverse is a virtual interactive online world, where users can interact with each other as well as with digital objects. It is considered the next phase of digital evolution that can revolutionize the digital adoption to a whole new level and extends the domain of services beyond the standard systems with online access and interaction capabilities. Digitization of services has become very popular for improving the efficiency in the fields of business, entertainment, education, e-trade and more. Due to the improved efficiency of digital services in those areas, their digitization has already progressed significantly. In a virtual digital interactive world, services and systems can be greatly enhanced and improved with the capabilities provided by digital systems, such as online storage, remote data processing, and cloud platforms. With the efficiency, performance, and the quality of service reaching their maximum potential, the metaverse perspective has skyrocketed along with the demand for an enhanced interactive experience.

The goal of this research proposal is to conduct a thorough study for the efficient resolution of combinatorial optimization problems.

This task will be accomplished by designing and implementing new optimization algorithms. The proposed doctoral thesis aims to design and implement parallel methodologies that will also use machine learning techniques to effectively solve computationally hard combinatorial optimization problems.

In the recent years, both traditional and new applications with challenging r equirements are being deployed to cloud computing environments. Such evolution is c haracterized by the communication of large data volumes over multiple network paths. C onsequently, the efficient traffic steering and compute resource management among network o r server nodes are core issues to be investigated, especially for data center (DC) networks and c loud environments. Load balancing in computer networks is an important technique sharing e fficiently the data traffic or workload over multiple network links or server clusters. A key f eature of cloud computing represents the ability to timely scale the computing resources a ccording to the application workloads, i.e., called elasticity. T he main aim of the present research proposal is to improve the applications’ p erformance in cloud computing environments, while utilizing the DC resources efficiently, t hrough applying novel load balancing and service elasticity techniques. We plan to study p roblems that in a general context are summarized in the following: a) efficient monitoring of network traffic that is adaptive to the features of specific applications (e.g. to identify b ottlenecks or unacceptable delays depending on the type of application) b) developing a m echanism to measure the degree of the application impact on server compute resources, c) the p roposal of a new load balancing mechanism, which has as an the novel aspect considers the t ype of application, d) the development of a mechanism that orchestrate micro-services for s calability purposes, e) the implementation of experimental scenarios for the evaluation of the n ew algorithms in relation to representative literature proposals.

Software quality and reliability are aspects of major importance for software development enterprises that wish to deliver high-quality and dependable software to their customers. Modern software applications are usually large, complex and interconnected through the internet and therefore they are constantly exposed to malicious individuals. Following secure and good programming practices, is necessary to develop secure software. The purpose of this study is to develop techniques that can provide indicative information to developers about the quality and security level of their software by identifying vulnerable hot-spots in the source code. Vulnerability prediction models are normally built based on machine learning techniques that use software attributes as input, to discriminate software components between vulnerable and clean components. This study aims to construct deep-learning models to predict whether a software component is vulnerable, utilizing the raw text of the source code in a sequential form, adopting techniques from the fields of natural language processing and text classification. Furthermore, during this study methods based both on text mining and software metrics as indicators for the existence of vulnerabilities will be considered. An investigation of the relation between technical debt metrics and the existence of vulnerabilities will take place, as well.

In the scope of the constantly changing nature of innovation and the continuous competition, the successful outcomes are mainly based on strategic knowledge sharing. The systematic transfer on know-how allows to create, transfer new values and discover innovative paths, that go beyond the existing processes and policies. This is applicable for example in modern enterprises, where for example all the established legacy domains that are an integral part of them, such as accounting are insufficient for an accurate assessment of the existing knowledge. As an outcome, there is a high risk of not considering knowledge elements, or even ignoring value their business value. Thus, a main challenge nowadays is to find the optimal combination for the main ingredients of the intellectual capital in an enterprise, as well as to determine the relationship between intellectual capital and the enterprise’s performance. Specifically, the intellectual capital has to be considered as added value for financial institutions. However, its contribution has not been consideration up to this moment in the Greek hotel business. Hence, we aim through our research to investigate the influence of intellectual capital on the performance of hotel business, but also to evaluate its contribution through the accounting and information systems perspective. During the basic steps of our work we have distinguished three important terms that characterize the Intellectual Capital was used, namely the human capital, the structural capital as well as relational / customer capital. These terms play a decisive role in the performance and evolution of an enterprise and are closely related. As a methodology for advancing with this work we relied on data derived by hotel executives, who replied to given questionnaires. Afterwards, with the use of statistical analysis along with the appropriate tools and methods, we have drawn conclusions that will be presented as a part of our final doctoral dissertation report. In this report we also demonstrate that the intellectual capital increases competitiveness and profit margin, enabling the use of cognitive knowledge and its quantification

The main research question of our study is to demystify the processes and stages that unfold within digital customer experience, in order tο:

i. Provide a seamless, stimulating, smart, synchronized and sensitive customer journey

ii. Ensure the business continuity of a given enterprise through a successful digital customer experience strategy

iii. Identify the mechanisms of the digital journey that cause the shift from the state of dependency to the state of empowerment for the digital customer

We propose a methodological construction named ‘’Digital Customer Continuity’’, which consists of the interaction and the evolvement of the digital customer in a bank through the channels and the added value, which is mainly created by 5 key components:

1. Digital Transformation

2. Digital Customer Experience

3. Digital Customer Engagement

4. Digital Customer Loyalty

5. Digital Customer Journey Mapping

The aforementioned components can be best visualized as a flow, or -even better- as a journey of the digital customer.

Marketing grows rapidly, through the adoption of technological innovations and the adaptation into the changes which are done in the social web. Programmatic Advertising plays an important role in Digital Marketing, as it manages to reflect the rapid changes in the Advertising Market which happen due to the use of modern information technologies.  The Programmatic Marketing promises to the advertisers the achievement of the 3-Rs, or in other words the ability to send the Right message, to the Right consumer, at the Right moment. However, Programmatic Marketing is not limited only in fulfilling the aforementioned. Using the unprecedented possibilities offered by the Artificial Intelligence and the Machine Learning, it attempts to understand the context in which human lives evolve and analyze the complexity of the real world. This new advertising ecosystem includes entities with distinct roles that interact with each other, digital platforms which make automated ad buying possible, and different transaction Models. Initially, the present research focuses on the systematic literature review in order to identify and analyze the recent and previous research approaches on this scientific field, as well as to record the existing literature based on certain criteria. Then, we aim to develop the  Programmatic Ecosystem with simultaneous reference and analysis of the entities involved and the procedures. The most important objective though is mapping the Programmatic Ecosystem and more specifically the Greek Programmatic Ecosystem. Subsequently, we proceed in the formation and the documentation of a research framework, based on the aforementioned, with a view to approach the Greek reality using empirical research methods. Finally, we attempt to design and conduct the mentioned research focusing on the Greek Market.     

Key Words: programmatic marketing, programmatic advertising, dsp platform, demand, targeting, impression, ad inventory, automation, consumer, data

Time-Driven Activity-Based Costing (TDABC) is considered by many researchers as a simple and useful costing system for library executives, who want to perform efficiently cost analysis. By implementing TDABC in Greek academic libraries, the management will focus on the costs of key activities. This, should lead to a better understanding of what causes costs, and, by extension, what changes are necessary to reduce costs

The framework of this dissertation, the application of econometric methods using mathematics and statistics for forecasting performancein terms of economic variable evolution is carried out. This variable is the 28 EU member state unemployment time series. The future forecasting unemployment rate is performed using three different methodologies, which are the main focus of this dissertation. The first section includes the Box-Jenkins methodology with the development and construction of ARIMA models as predicting tools of economic variable values. The Box-Jenkins approach in time series analysis is a method of finding an ARIMA model that adequately represents the stochastic process from which the sample originated and this ARIMA model enables unemployment forecastingfor the countries being investigated. In terms of forecasting, a methodology widely used in economic or econometric models is the Kalman filter, which is the second section of this dissertation and is considered as one of the most commonly used methods for estimating, optimizing and predicting data. The next partof this dissertation concerns forecasting with the use of neural networks;a computational technologyto predict economic variables. This technology is used in time series forecasting, in order to predict future values by way of appropriate neural network training. The aim of this study is to measure the accuracy and effectiveness of each forecasting method. This study concludes with a comparative assessment of the alternative techniques of each method via a common criterion, which will identify the optimal method in the prediction process. Forecasts are widely used in both the public and private sectors for policy and decision-making. The application of forecasting methods includes two steps; analyzing data series and selecting the forecasting method which best fits the data in order to safely predict the future

Human aspects of Software Engineering is a research field focusing on the locomotive of the software development process, the engineers themselves. It is a steadily growing field since as the software industry matures, the more it is accepted by the software engineering community that people involved in software development processes deserve more attention than the processes or technologies themselves. The importance of these issues is widely acknowledged in the industry and academia, because many of the failures of software systems can be traced to human factors. [1]. The aim of this dissertation is to examine the effect that the plethora of programming languages and frameworks have on the engineers (from the cognitive and learning perspective but also from the perspective of the skills that have to be acquired), thereby investigating the impact on project development and maintenance.  Our goal is to undertake quantitative studies on the way the new languages and frameworks are affecting developers and projects over time by mining multiparadigm projects at large scale from software repositories and then create tools to produce visualizations and provide insights of this relation.

This PhD research, is concerned with the development of an Augmented Reality environment capable to transmit motor skills. To achieve this goal it combines various algorithms to implement a methodology that exploits motion capture technologies to accomplish the  comparison of the gestures executed by a learner with the reference gestures of an expert.

This research concerns the scientific area of e-learning systems or environments.
This particular type of education/learning has gained great importance in the modern,
advanced technological landscape, as a large part of education today is delivered in elearning
environments where learners engage in a variety of learning experiences. The
quality of education is a very important parameter for the sustainable development of a
state, while it is necessary for every educational institution to have a clear policy regarding 

the assurance of quality in learning and particularly in e-learning. Analyses show that quality
will define the future of e-learning, so two key questions arise: what is quality (of e-learning)
and how can we ensure it? It is of major importance to monitor and evaluate the quality of
learning experiences in the various e-learning environments because it affects the learning
outcomes of learners. This research deals with the theme of evaluating the quality of
experience in e-learning. The factors affecting the quality of experience will be investigated
and a model for its evaluation will be designed. It will also be studied whether the transition
to advanced VR/AR environments increases the quality of experience of learners.

Digital identities are an indispensable component of modern information and communication systems. Their influence spans the technological, social, and economic dimensions of cyberspace. In essence, during the last decade, there has been a paradigm shift concerning how identity information is defined, presented, updated, and deleted.

 This research proposal addresses the emerging approaches to the field of identity and access management (IAM). Among others, it emphasizes the decentralized approaches to identity management, advanced methods for access control, and modern mechanisms for authentication.

 In this direction, the research proposal sets out three objectives: a) to determine and evaluate the current developments in the area of identity and access management, b) to contribute to the field of identity and access management in terms of information security, data protection, and privacy, and c) to contribute to the field of identity and access management regarding the support of additional use-cases.

Nowadays, we have to do everything we can to improve the nature of instruction. Currently, innovations such as Artificial Intelligence (AI), Virtual Reality (VR), Internet of Things (IoT), Blockchain (BC), and wearable devices are being used to improve the education sector's ability. The technology of smart classroom by which AI applied in the instruction, is called Smart Classroom. The IoT is rapidly expanding and the number of devices we will be interacting with on a daily basis is tremendous. Artificial Intelligence of Things (AloT) uses artificial intelligence to process the data from sensors and send it to the application layer for decision making. Also, BC is an innovation to provide security in transactions between IoT education users as students, teachers, employers, developers, facilitators, recruiters, and accreditors on the Internet. Also, Digital Twin technology is Known as one of the key technologies for building a new online learning platform , which will play a vital role in the field of education. These technologies will become an extremely important part of academies around the world shortly. The smart classrooms are not too far away. This research will investigate security in the wider ΙοΤ using Al in 6G networks in the field of education, with the aim of identifying their weaknesses and security gaps, which may there are.

In recent years, the field of computer science has increasingly recognized the importance of writing
high-quality code. Both the industry and the research community have made it a priority to maintain
good quality in their codebases. Preserving good quality in a project positively impacts its lifecycle by
enabling easier understanding, reusing, extending, and debugging the software. One of the key
indicators of code quality is Technical Debt, which lies at the core of research on software quality and
technology in the last two decades. This doctoral thesis aims to explore methods for an automated
assessment of Technical Debt, as well as new factors that are correlated with its accumulation rate. By
utilizing these correlations, we will experiment with predicting the evolution of a project’s quality using
state-of-the-art artificial intelligence methods. As a result, we will be able to assist in managing
Technical Debt through its timely detection and contribute to the overall improvement of the quality of
complex software.

Serious game-based education is gaining more and more supporters day by day. Serious games have been shown to be effective in educating and training in a variety of fields including cyber security. It is generally accepted that collaboration between  learners plays an important role in education. Although cooperative serious games have been shown to improve trainees' motivation and learning outcomes, only a limited number of games are designed to promote collaborative learning. The purpose of this research study is to investigate the existing design and development models of serious games, with an emphasis on those that are oriented towards collaborative learning. The sub-objectives are: a) the identification of the factors that could affect the effectiveness of learning and education based on serious games, b) the proposal of a model for the development and design of collaborative serious games and c) the design and development of a collaborative serious game for cybersecurity learning and training.

Time series forecasting is a critical task in many fields, with applications ranging from finance and economics tometeorology
and engineering. In recent years, deep neural networks have been shown to be particularly powerful for time seriesforecasting,
achieving state-of-the-art results on a variety of tasks. However, there has been a trend towards relying too heavily ondeep
learning techniques, known as black boxes with great results, with little consideration of classical explainablestatistical models.
This proposal aims to address this issue by trying to infuse deep learning models for time series forecasting with theexplainability
of classical statistical models.

Recently, the term "haptic" comes to the surface by referring to technological applications related to the science of manipulating and identifying specific objects through touch. During the PhD thesis, emphasis will be placed on exploiting the capabilities of 5th generation (5G) networks in order to exchange haptic data without interruptions. In addition, the functionality of these two technologies will be studied in order to examine the characteristics that will lead to the development of innovative applications that will contribute to improve today's needs. The main objective of this research will be to combine the advantages of 5G wireless networks and the capabilities that haptic communication can offer in order to examine the correct sending of such type of data and to discover the benefits that humanity can obtain, through the implementation of innovative applications. This could be done by studying and developing new wireless communication protocols that can be used for the purpose of improving communication between the transmitter and the receiver. In addition, the security challenges of combining these technologies will be examined.

The goal of this thesis is the study of the evolution and management of Technical Debt in modern software projects. High competition and short delivery deadlines have made Technical Debt a major software development problem. This problem cannot be entirely solved, instead every effort to resolve it focuses on limiting it to a sustainable extend. The Technical Debt is generated during the entire life cycle of a software project, starting from the requirements analysis and design to the final delivery to the customer. It is important for this study to include the different stages in the software development process without being limited to code quality. The next step is to research the several tools that measure technical debt and the methodologies that are used to calculate it, then to evaluate each tool comparatively and to determine their degree of agreement. It is important to include in the equation the concept of interest so that the overall impact picture is as representative as possible. The aim is to contribute with this thesis to the effort of building a common ground for studies in the field of Technical Debt.

The analysis of the financial statements of a financial unit is one of the most important parts of Financial Accounting. Its implementation provides all kind of information to the financial institution that applies it as well as to the rest of the business world. The financial statements provide the necessary information to the parties involved in order to take the right decisions regarding the economic units. Through the use and correct interpretation of the financial statements and with the contribution of methods of analysis, we can obtain information about the course of the financial unit and its financial position. The most effective and the best source of information on an enterprise's economic activity is the balance sheet, the income statement and the statement of cash flows. These financial statements analyze the assets of the enterprise and the sources of its capital. In order to better understand the usefulness of the analysis of financial statements, we will examine a number of various financial ratios, which are considered to be crucial to making good business decisions. We will then proceed in practical application by choosing a business in which we will implement and analyze these financial ratios where we think that are important and necessary to be analyzed for the benefit of all concerned. In addition, a section of this research will deal with the cash flow ratios as well as with the bankruptcy prediction, through the use and analysis of financial statements. Finally with the help of Information Technology will take place selection, collection and processing of information mainly by using Internet tools, spreadsheets, charts, Pivot spreadsheet data, graphs, trends and sizes comparisons in order to get the best possible analysis of the financial statements and in particular the Cash Flow Statement.

In recent years there has been a great increase in interest in technical debt management, both by industry and researchers, due to the positive contribution of proper debt management. Despite the decades that passed since this term appeared in the aspects of software technology, several questions remain regarding its proper management. In addition, as new software development methodologies are introduced for modern applications, many technical debt management activities will need to be reviewed. In this dissertation, factors related to the rate of technical debt accumulation in software developed with DevOps methodologies will be studied. Moreover, the procedures that are being used for the management of the technical debt in modern software applications will be studied, with emphasis on quantification. Therefore, we will be able to shed some light on the management of technical debt in modern software applications and modern ways of development.

For centrally controlled business process execution, various important concepts such as conformance, reliability, trust, and transparency, have been over time a roadblock. Without global monitoring, participants have restricted visibility of the overall progress, thereby establishing a basis for misunderstandings and shifting blame in cases of conflict. Emerging blockchain technology has the potential to drastically change the environment in which processes can operate. The underlying doctoral research aims to identify where the intersection of business processes and blockchain lays, as well as to propose a modeling framework, utilizing the BPMN process modeling language, for the graphical representation of blockchain-based processes. As a result, the integration of business processes with the blockchain technology will be meticulously rendered in a standard way, before being translated into smart contracts code and enable the development of Decentralized Applications (DApps) atop blockchain.

This dissertation aims at the utilization, integration and further development of methods and algorithms from Artificial Intelligence, Game Theory, Information Theory and relevant frameworks for exploring new methods, applications and tools for decision-making. More specifically, this research aims to broaden the understanding of topics of cognitive-scientific interest in general through simulations as well as the applications of Intelligent Systems. Key objective is the extraction of knowledge and the establishment of innovative approaches for the field of Artificial Intelligence and Information Theory. Also, in the context of this doctoral dissertation, the introduction, modeling, merge, comparative analysis and application of existing and new efficient solutions and methods in new and / or existing problems and environments are launched. The aforementioned aim to of draw conclusions, study and optimize algorithms under various conditions, as well as to highlight benchmarks as springboards for further studies of  interdisciplinary interest.

Nowadays, and especially after the pandemic, the labour market has changed. Employers are looking for people that have not only the technical knowledge but also a sophisticated skillset. All the terms related to skills, soft or not, digital and entrepreneurial are in the center of discussions of employers and employees.

The majority of people who search for a new job are highly educated and in a constant search for new courses to follow so as to become even better in their job.

On the other hand, human resources departments are looking for leaders, people that can work strategically, can manage teams and projects, but also professionals in the IT and all the new technologies that constantly arise.

What is the skillset that someone should home in order to address the needs of the new jobs to come in the area of informatics? In which categories they can be clustered and how do they impact the effectiveness of the employer? This research will try to to answer these questions by clarifying the skills needed for the different new jobs to arise in the area of informatics and also make a matching between the skills and these jobs.

From 2010s, ports and terminals have evolved towards a ‘smart port’ concept.  In the port operations and management research field, performance measurement has been one of the main issues. In addition, the concept of efficiency is vague and proves difficult to apply in a typical port organization extending across production, trading and service industries. Given the fact that the evaluation of the performance of such a complex system includes the measurement of conflicted criteria and especially the measurement of the importance of such criteria by a group of experts, it is verified that Multi Criteria Decision Making (MCDM) methods should be applied so as to evaluate the overall performance of the examined system.

Only few researchers have developed a conceptual framework of a smart port while the global industry has been driven under 4th innovation industry. Furthermore, even fewer studies have focused on applying fuzzy logic-based methods to evaluate the performance/efficiency of smart ports in a multidimensional concept applying to the “industry 4.0” terminology. Under this assumption, our study contributes to the literature in a twofold way: first of all, our aim is to define the concept of a smart port, as well as the main elements that constitute a port "smart" and also to provide smart port performance dimensions and related criteria in a new perspective by proposing a hybrid fuzzy conceptual model which claims its novelty on assessing overall performance of smart ports, including an approach that takes into consideration the weighing of specific indicators that groups them into clusters of interrelated dimensions.

In the "big data" era, data generated digitally on a daily basis is expanding at a rapid pace with the
advent of technology. However, conventional analytical methods and techniques may not be
sufficient to analyze and extract insights from such massive datasets due to the rising volume and
diversity of data being generated. Big data analytics approaches handle enormous amounts of data
from various sources so as to uncover patterns, trends, and correlations that could guide decisionmaking
and make valuable predictions. Moreover, big data sentiment analysis may assist a
corporation in gaining meaningful insights from text-based content. At the same time, there is a
rising academic interest in AI and, more specifically, in applying machine learning approaches and
predictions through technological breakthroughs. Machine learning approaches are referred to as
being beneficial in situations involving vast volumes of data. AI big data analytics is the process of
analyzing and extracting insights from massive and complex datasets, utilizing artificial intelligence
tools. Big data analytics can apply AI methods, such as machine learning approaches, to uncover
hidden relationships and patterns and to automate and improve the analysis process. The main
objective of this doctoral thesis is to generate accurate predictions utilizing machine learning as well
as sentiment analysis and provide a robust and adaptable data analysis technique against the
challenges.

In today's digital age, developments in Machine Learning, Natural Language Processing, Robotic Process Automation and Deep Learning technologies are leading more and more companies to turn to Artificial Intelligence to make their operations smarter and more efficient. On the contrary, as business research of Artificial Intelligence from a strategic point of view is at a fairly early stage, results in its limited utilization. Therefore, many companies, without realizing the critical factors that lead to an Artificial Intelligence Capability, make wrong decisions or make poor investments in such technologies. Consequently, they fail to improve or discover new organizational capabilities that make them gain a competitive edge and stand out in the digital transformation of today's business world. For this reason, this dissertation aims to map the relationship of Artificial Intelligence to Digital Strategy and to explore the factors that lead through it to value creation and improved business performance, in order to enable the proper design, use and evaluation of various digital strategies, oriented to the highly competitive conditions of modern digital era. The results of the present study are called to provide understanding of both the internal and external factors of a modern business, as well as to identify the mix of natural, human and organizational resources required to create an Artificial Intelligence Capability, which can offer significant value and bring increased benefits to overall performance.

This research proposal attempts to identify and focus on the key legal issues that arise from the interaction of Legal Science with Digital Forensics and then to determine if there are cases where legislation can protectively surround the challenges faced by the field of Digital Forensics due to the rapid development of technology. This is a theoretically attractive field of research because of its inherent interdisciplinarity, in the sense that the examination of digital evidence, which was created more as a necessity driven by the rapid technological developments of recent decades and less as a novelty, is the application of scientific methods and tools to digital data to be used as evidence in investigations and legal proceedings. At the same time, however, it is a practically important field because Digital Forensics, although assisted and supported by the science of Informatics, is very closely defined by legal requirements. Moreover, its development, progress, and efficiency depend on and are guided equally from legislation and case law, both at the national and European level.

Over the years, eGovernment is growing rapidly, reaching the point of being a ‘global issue’. eGovernment refers to government services based on the internet and one important focus point is the provision of information to citizens/businesses about public services. This information provision can be significantly improved by employing emerging technologies. Chatbots are already used in the private sector offering advantages such as availability 24/7, simultaneous service of many customers, automatic learning and updating, etc. Knowledge Graphs are also used in the private sector and offer advantages such as scalability, ease of understanding, easy querying, etc. Opportunities and challenges arising from the exploitation of these technologies are many. To the best of our knowledge, no relevant research exists about using Chatbots and Knowledge Graphs for eGovernment information provision based on EU standards. With this doctoral dissertation we aim to fill this gap. To accomplish our goals, we will employ the design science methodology in order to develop a holistic framework (including architecture, pilot applications and a roadmap) for public authorities to integrate chatbots and knowledge graphs using the European data model CPSV-AP for information provision to citizens/businesses about public services.

In recent decades, a large number of developing and developed economies have faced some form of banking crisis, such as the bankruptcy of a bank. Banking crises disrupt the normal function of the financial system and provoke negative consequences for the economy. Predicting a potential banking crisis is a matter of increasing interest for the supervisors and the financial institutions in order to ensure the stability of the banking system. So, it is vital to develop tools that can identify an upcoming banking crisis. This doctoral thesis aims to predict the factors that lead to the bankruptcy of a banking institution using machine learning techniques. In order to achieve this, a large sample of American and European banks that have been bankrupt for over a period of 15 years will be used. In addition, the use of both American and European banks will identify the differences and similarities that lead to a bank financial failure in each region.

Learning the violin has for decades posed several difficulties for students of all educational levels. The great time needed for novice violin musicians to be able to master the essentials of the violin is an indicator of this music organ's particularities. The research community has presented evidence of benefits by incorporating serious games in the learning process. Since there are some computer-assisted violin training systems (CAMIT systems), which are not provided under the scope of a serious game, there is a need to analyze today's serious games that focus on learning musical instruments. The main objective of this research is to design, implement and evaluate an application that manifests all the characteristics that constitute a comprehensive serious game for learning the violin and contribute to enhance students' learning outcomes, increase their performance as well as motivate continued learning. With the results of conducting this research, we look forward to integrate the game into music schools in order to test and evaluate it. The feedback from the students will be used to observe the degree of its effectiveness and to trigger future research into serious games related to violin education.

The difference between the traditional network structure and the newer one, the
heterogeneity of the networks, the complexity, and the interoperability they
provide, determine the different models and management methods to achieve the
highest possible network security. Computer systems and modern integrated
networking platforms, usually, consist of many nodes and make use of advanced
hardware and specialized software through actions that are as automated as
possible.
Including technical measures, the implementation of administrative measures is
necessary. The new requirements of modern and intelligent networking require
specialized frameworks, management optimizing ways and threat response
strategies, which offer the ability to control all the managed components of a
network system, the resources and services, the ability to monitor proper
operation, and policy - based management capabilities. A common requirement for
all modern networks is to protect them from inappropriate use and to avoid
violations against any intruder.
The contribution of this dissertation, having aware of the current situation, is
to develop training models and methods for the management and security of networks
systems, as also to present an intelligent analysis and management system with
an emphasis on dealing with security incidents that can be adapted to different
approaches.

With the radical improvement of technology, the availability of large databases, and the increased computational power, Machine Learning (ML) algorithms have achieved excellent performance on complex tasks in multiple application domains. Specifically, they play a key role in Industry 4.0 by powering intelligent machines to execute tasks autonomously such as self-monitoring, interpretation, diagnosis/detection, and analysis. However, ML-based models, most notably neural networks that offer the greatest potential, are hindered by the lack of interpretability, since their operations are opaque to the human users. As it is difficult to generate value for business-to-business (B2B) marketing operations from opaque models, in recent years, AI researchers have tried to make black-box models transparent or “glass box” through AI explainability (xAI) approaches. Our purpose is to optimize the operation of industrial machinery and present the solution to human users in an understandable way, in order to facilitate the decision making process.

In the field of computer science, most of the serious games aim to teach programming. Despite the interest of academic and educational institutions as well as the innate predisposition of students for play, their use in schools is limited. Many serious games fail to take full advantage of their strengths and keep students interested, discouraging their use. On the other hand, the digital entertainment industry is well established and has been successfully implementing digital game learning design techniques for its users for years. By studying these techniques, the proposed thesis will help to explore methods of designing educational activities as part of the mechanisms of the game, promoting authentic learning and enhancing a meaningful learning context.

 The aim of this thesis is to research those elements of commercial games that motivate user engagement and can be incorporated into educational programming games. It will be examined whether the integration of educational activities in the mechanisms of the game brings about changes in the learning outcomes. Finally, it will investigate whether individual teaching of basic programming structures can be simulated by dynamically creating an adapted level of difficulty educational scenarios.

Peer-to-Peer (P2P) lending has seen a significant development through the vast amount of data recognition and the revolution of FinTech. P2P lending platforms which are regulating the relationship between borrowers and lenders are still sensitive to multiple types of risk management. In that case, different methods are applied to mitigate risks at P2P lending platforms. Conventional machine learning algorithms suggest high forecasting performance, but many of them lack interpretation and explainability. This PhD thesis will focus on a two-stage hybrid credit risk prediction model based on LightGBM and graph-based deep neural networks. This model is going to compare with different machine learning approaches involving single classifiers, heterogeneous ensembles and graph learning. Moreover, advanced post-hoc model agnostic eXplainability AI (XAI) techniques, are used to evaluate ML-based credit scoring techniques based on the P2P Lending Platform. Furthermore, implementing machine learning develops new investment models that can maximize profitability while minimizing risk, local and global instance models, based more on visualization for better interpretability.

One of the most pressing issues in production today is increased complexity. Although product variants are the most common cause of complexity in a production system, other factors have been considered, such as product structure, plant/shop structure, planning and scheduling, information flow, dynamic variability and uncertainty of the environment, and information and material flows. For managing the production complexity, methods and techniques developed in different disciplines have been tested. For example, genetic algorithms have been used in manufacturing systems for order sequencing and production scheduling, and neural networks have been used in decision support. This PhD thesis aims at examining several production algorithms and their optimization. Extensive research in the parallelization of production algorithms will lead us to the implementation of a new parallel production algorithm or a suggestion of certain improvements on the state-of-the-art algorithms by employing parallelism techniques.

Knowledge Graphs are an essential area among current trends in technology. A Knowledge Graph is practically a method of representing information within a knowledge domain in the form of relations between entities of this domain. Additionally, by adopting Semantic Web and Natural language Processing technology, Knowledge Graphs have become a powerful tool not only for accessing information effectively but also for expanding knowledge by using inference mechanisms. However, despite the significant technological development in the field, there are still several open challenges related to generating a knowledge graph from some data source, aligning different knowledge graphs or inferring knowledge. In this study, we focus on domain-specific knowledge graphs and particularly on aligning graphs which represent knowledge of similar domains, aiming at the construction of a more complete knowledge graph corresponding to an extended-domain, which will allow us to infer new knowledge.

A Smart City uses Information and Communication Technology (ICT) to improve operational efficiency and provide quality services by collecting data on its infrastructure, processes, and operations, such as utilities, mobility, and infrastructure. The necessary tools to successfully implement the "smart" city model are Internet of Things (IoT) sensors and IoT analytics platforms. The produced data contribute to managing resources, assets, and services and improving operations. Nowadays, two core technologies are discussed most about "smart" cities: the concept of the Industrial Internet of Things or IoT 4.0 and the Digital Twin. IoT 4.0 (or Industrial Internet of Things) represents a future upgrade of the smart city system segments. Integrating ICT and computing technologies enable cities to share data, information, and instructions between all agents and systems. The Digital Twin is a virtual representation of real-world objects, processes, behaviors, and relationships, is updated from real-time data, and uses simulation, machine learning, and reasoning to help decisionmaking. The research investigates the concept of Smart Cities, IoT 4.0, and Digital Twin and how they can be best implemented and combined.

The analysis of the financial statements of a financial unit is one of the most important parts of Financial Accounting. Its implementation provides all kind of information to the financial institution that applies it as well as to the rest of the business world. The financial statements provide the necessary information to the parties involved in order to take the right decisions regarding the economic units. Through the use and correct interpretation of the financial statements and with the contribution of methods of analysis, we can obtain information about the course of the financial unit and its financial position. The most effective and the best source of information on an enterprise's economic activity is the balance sheet, the income statement and the statement of cash flows. These financial statements analyze the assets of the enterprise and the sources of its capital. In order to better understand the usefulness of the analysis of financial statements, we will examine a number of various financial ratios, which are considered to be crucial to making good business decisions. We will then proceed in practical application by choosing a business in which we will implement and analyze these financial ratios where we think that are important and necessary to be analyzed for the benefit of all concerned. In addition, a section of this research will deal with the cash flow ratios as well as with the bankruptcy prediction, through the use and analysis of financial statements. Finally with the help of Information Technology will take place selection, collection and processing of information mainly by using Internet tools, spreadsheets, charts, Pivot spreadsheet data, graphs, trends and sizes comparisons in order to get the best possible analysis of the financial statements and in particular the Cash Flow Statement.

Artificial Intelligence has been proven to hugely benefit from big data and enormous computer power. This fact gives a great head start to the few large companies that possess such infrastructure, so they tend to dominate the AI space. A desirable alternative could be the creation of prototypes that: 1) No longer rely on large private data centers. 2) They are not based on just a decentralized model, but on a fully distributed one, meaning that they have no need for a central coordinator. Following the above directives, we could develop networks that: 1) Can approach (or even exceed) the total computing power of a big data center, by utilizing a large number of nodes. 2) Have no “single point of failure”, since all participants have equal roles, meaning that the network will keep working even if a portion of it gets offline. 3) Are democratized, since all nodes participate to the cause and can potentially have the exact same access to the produced outcome. So far almost all of the implementations regarding neural network training, focus mainly on performance issues. Although there has been significant research on distributed algorithms, there is not enough research on a model that has no need of a coordinating server. Our proposal utilizes new peer-to-peer topologies based on cryptographic verification (Distributed Ledger Technology – DLT), attempting to diffuse the server role to the whole network. During our research we will test various configurations of the distributed model, in order to evaluate their impact on both the training speed and the quality of the produced model.

This PhD dissertation explores the strategy of exploiting real-time user location-based data and studies the pronunciation of targeted products and services organizations through the development of a complete framework that will increase their competitiveness

The PhD dissertation focuses on the digital transformation and how this strategic decision affects the adoption of digital technologies, customer experience and businesses’ operations.

Cryptoeconomics are characterized as complex systems that combine various principles, theories, methods and techniques from the sciences of computers, mathematics and economics. They are based on studying economic incentives and utilizing the Game Theory. At the same time, Cryptography and Distributed Consensus are used to develop reliable decentralized applications. In this context, our research aims at analyzing and modeling the use of Cryptoeconomics for the benefit of Cybersecurity.

The Dissertation explores the possibilities of simulating the phenomenon of demand for goods, as a significant economic activity by households, by creating software based on a model of simulation of this phenomenon.

The transfer of personal financial data as a phenomenon of modern technological and economic reality raises issues whose investigation requires the compilation of both Computer Science and Legal Science. The evolution of digital applications and hence the possibilities offered to individuals and other entities has made it necessary to provide specialized legal protection. The scientific scope of this dissertation follows the rapid legislative evolution of the European Union by the adoption of laws relating to the protection of personal data from unauthorized processing by individuals and by economic entities. The implementation of the new international institutional framework in the area of ​​personal data, with the transmission of a huge amount of personal data nowadays by international internet companies (facebook, google, apple, microsoft, yahoo, etc.) requires the research in both the technological environment, economic reality and the accomplished institutional framework in response to emerging scientific concerns. Presenting, analyzing and formulating proposals on the legal framework of financial personal data is intended to create a useful manual for clarifying the transmission of personal data.

The dissertation focuses on the automated modeling of information related to "cyber threats" and can be described by standards that fit into Cyber Threat Intelligence, in order to make it useful in advanced security applications. The main research objectives of the dissertation include the following: • Development of knowledge representation models, derived from the analysis and study of cyber threats. • The representation of information that reflects the observable cyber threat activity and the development of methods and techniques of automatic creation of artifacts. • The development of techniques to optimize the extraction of information about cyber threats from public repositories. • The development of automated information technology techniques in the field of Cyber Threat Intelligence.

As the leading writer and Nobel laureate George Bernard Shaw (1856-1950) said, “Money is the most important thing in the world. It represents health, strength, honor, generosity and beauty as conspicuously and undeniably as the want of it represents illness, weakness, disgrace, meanness and ugliness.” (Walsh, 2006). In the present study, the main focus of the study will be to analyze a company's financial statements with the help of indicators in order to present the usefulness and the amount of information provided by these financial statements. A special mention will be on Cash Flow statement because, with the introduction of international accounting standards, the presentation of the cash flow statement became an integral part of the financial statements (Leonie Jooste, 2006). Although it has many enthusiastic supporters since its introduction in 1986, is observed that there is not been written or developed many theories, for the effective use or analysis of the Cash Flow Statement (Carslaw & Mills, 1991), while several authors agree on the importance and usefulness of the Cash Flow Statement in Financial Analysis (Leonie Jooste, 2006), as it emerges through its analysis the solvency and the ability of the firm to meet its short-term liabilities. (Carslaw and Mills, 1991). The purpose of this research is in-depth study of the International Accounting Standards 7 Cash Flows and its vital role in the proper operation of an enterprise as well as the other financial statements of an enterprise. With the help of indicators and the use of information systems, we will give a clear picture of results that will be radiate from the actual published financial statements of Greek companies all operating in the iron production and trading sector for the best possible analysis of financial statements and especially the situation Cash flows.

Business process management is a discipline which seeks to increase the adaptability and value creation potential of companies by holistically analyzing and optimizing their processes regardless of departmental boundaries. But, in order to be able to analyze a process and to tap the full efficiency potential a thorough understanding of the process is required first. The necessary level of understanding can be achieved by creating formal models of business processes. Such business process models act as a means to document and analyze the underlying process and can serve as a blue print for subsequent implementation and automation activities. Several formal or semi-formal graphical notations are available to map a process to its graphical representation containing the performed tasks, involved actors, required data and documents, and business rules, which describes the logical and temporal aspects of the process execution. Ko defines business process as a series or network of value-added activities, performed by their relevant roles or collaborators, to purposefully achieve the common business goal (2009). Natural Language Processing (NLP) is an interdisciplinary field which combines artificial intelligence and formal linguistics. NLP is concerned with analyzing and understanding human language both in written and oral form. Human languages are ambiguous in many levels, thus creating applications that can comprehend them is very complex. However basic NLP tasks have been developed that can deal with ambiguity with high accuracy. Such tasks are tokenization and part-of-speech tagging and they form the building blocks of almost any NLP application. The aim is to explore how NLP can be applied in the BPM domain in order to automatically generate business process models from existing documentation within an organization. The main idea is that from the syntactic and grammatical structure of a sentence the components of a business process model can be derived (i.e. activities, resources, tasks, patterns). The result would be a business process model depicted using BPMN – a dedicated business process modeling technique. The extraction can be achieved by applying well known NLP techniques such as Part of Speech tagging, Named Entity Recognition and Co-reference Resolution. The ultimate goal is the proposal of a system which is able to generate conceptual models needed for a BPM-initiative from text. Using our approach a system analyst is relieved from the time-intensive modelling task. By pursuing this research goal, we evaluated related works in the field of BPM and NLP, developed a novel transformation approach, and created and evaluated it.

This dissertation explores the creation of a sensorimotor learning system for teaching microtonal music, which aims at the recording and sound (melody) recognition along with the corresponding gestures of the expert. The system will provide feedback as well as visual learning analytics

Business process optimisation is the automated improvement of business processes using pre-specified measures of performance (criteria). The importance of business process optimisation lies in its ability to (re-)design a business process based on quantitative evaluation criteria. This concept stresses the need to generate alternative business process designs based on the given process requirements, and quantitatively evaluate and compare these designs. The intended process optimisation can be performed through Evolutionary Computing (EC) techniques, as they have been successfully applied to several combinatorial problems and demonstrate a series of benefits. One significant advantage lies in the gain of flexibility and adaptability to the task in hand, in combination with robust performance and global search characteristics. To assess the capabilities of the proposed framework and the extent that these contribute to business process research, the performance of the framework and the generated business process designs need to be quantitatively measured and evaluated. The proposed research involves response experiments that investigate the effect of several variables at different levels. Therefore, it requires an experimental strategy to be undertaken.

The dissertation proposes a new method for the detection of nonlinear long term and short term relationships of co-movement, between Macroeconomic and Financial variables, widely known as Cointegration testing. We expand the theory of the current methodology, for the detection of Cointegration, testing and modeling linearized long and short term relationship between the economic variables, when the relationships are characterized by nonlinearity. For the linearization of the nonlinear relationships between the economic variables we consider classical linearization techniques (i.e. Gauss Seidel, Box-Cox transformations) and their extension to Threshold models. To evaluate the effectiveness of the proposed Nonlinear Cointegration method we use stochastic simulation techniques and then apply it to a number of selected Macroeconomic and Financial variables of the Greek and European Economy

Accounting science has been developed in order to provide the accurate information to users of the financial statements and is based on certain basic principles, compliance with which is a basic prerequisite for the preparation of sound and acceptable financial statements. Accounting science follows business development, which creates the need to study new conditions and to create new accounting standards. The new standards should be objective, valid and complete, always based on the standard principles and rules. According to the introductory report, the implementation of HELLENIC ACCOUNTING STANDARDS aims at the consolidation, completion and modernization of the accounting rules used in our country, in order to create a comprehensive and operational accounting and regulatory framework for the enterprises and other entities. In addition, their implementation fights accounting multiplicity and serves the need for transparency, reliability and comparability of financial information in all European Union enterprises, creating conditions which contribute to the smooth operation of the market. In the present research proposal, is being studied if the “Hellenic Accounting Standards” serve the above purposes, since they provide different types of information than the previous prepared financial statements and also offer different accounting treatment for specific cases and events. With regard to the different types of information, it is indicated that additional information is now provided on the following: Transactions with affiliated companies, The difference in valuation methods of inventories when using LIFO, Leased equipment and the time separation of future payments, Data on provisions and possible liabilities. While, as it concerns different accounting treatment, the following will be mentioned: Research and development costs, Fixed assets and depreciation, Participating interest, Provisions, Taxes for future years

The aim is to propose new methods and means for the development of digital serious games in Cultural Heritage. The results are expected to contribute to the cultural heritage serious games’ efficiency in terms of their learning outcomes and communication of the cultural content.