No description available.

In recent years, a number of laws and regulations (such as the Basel II accord or SOX) demand that organizations record certain activities or decisions to fulfill legally enforced reporting duties. Most of these regulations have a direct impact on the information systems that support an organization's business processes. Therefore, the definition of audit requirements at the modeling-level is an important prerequisite for the thorough implementation and enforcement of corresponding policies in a software system. In this paper, we present a UML extension for the specification of audit properties. The extension is generic and can be applied to a wide variety of UML elements. In a model-driven development (MDD) approach, our extension can be used to generate corresponding audit rules via model transformations. (author's abstract)

We have developed a tool that can create a Java code generator for a behavioural modelling notation given only a description of the notation's semantics as a set of parameters. This description is based on template semantics, which has previously been used to describe a wide variety of notations. As a result, we have a technique for generating Java code for models written in any notation describable in template semantics. Since template semantics allows for models containing nondeterminism, we introduce mechanisms for eliminating this nondeterminism when generating code. We describe Java implementations of several template-semantics composition operators that have no natural Java representations and give some techniques for optimizing the generated code without sacrificing correctness. The efficiency of our generated code is comparable to that of commercial notation-specific code generators.

We have developed a tool that can create a Java code generator for a behavioural modelling notation given only a description of the notation's semantics as a set of parameters. This description is based on template semantics, which has previously been used to describe a wide variety of notations. As a result, we have a technique for generating Java code for models written in any notation describable in template semantics. Since template semantics allows for models containing nondeterminism, we introduce mechanisms for eliminating this nondeterminism when generating code. We describe Java implementations of several template-semantics composition operators that have no natural Java representations and give some techniques for optimizing the generated code without sacrificing correctness. The efficiency of our generated code is comparable to that of commercial notation-specific code generators.

PhD Thesis (Software Engineering) / Model-Based Development (MBD) is an approach that uses software models to describe the behaviour of embedded software and cyber-physical systems. MBD has become an increasingly prevalent paradigm, with Simulink by MathWorks being the most widely used MBD platform for control software. Unlike textual programming languages, visual languages for MBD such as Simulink use block diagrams as their syntax. Thus, some software engineering principles created for textual languages are not easily adapted to this graphical notation or have not yet been supported. A software engineering principle that is not readily supported in Simulink is the modularization of systems using information hiding. As with all software artifacts, Simulink models must be constantly maintained and are subject to evolution over their lifetime. This principle hides likely changes, thus enabling the design to be robust with respect to future changes. In this thesis, we perform repository mining on an industry change management system of Simulink models to understand how they are likely to change. Then, we explore the various modelling mechanisms available in the Simulink language to see how they could support modular design with information hiding. Next, we propose a module structure, syntactic interface, and modelling conventions for Simulink designs, which are supported by our open-source Simulink Module Tool. Finally, we apply the proposed techniques on case studies from the aerospace and nuclear domains, in order to demonstrate their practicality and validate their effectiveness. Overall, the approach helped support information hiding by encapsulating secrets and facilitating likely changes. It also had a positive effect on interface complexity, cohesion, and coupling. The larger system also exhibited reductions to cyclomatic complexity, testing effort, and execution time, but the smaller case study benefited less in these areas. / Thesis / Doctor of Philosophy (PhD)

In this thesis, we present an abstract meta-model for model driven development of web applications targeting multiple platforms. We review the existing technologies and the related work in order to obtain a list of requirements for such an abstract model. The abstract model is built by extending an existing UML-based model for web applications. We demonstrate that it is possible to map this abstract model to more than one specific development platform by providing transformations for these mappings. We also lay out the general outline of a model-driven process based on the proposed abstract model. The abstract model and the model-driven process are supported by a set of tools, case studies and a visual modeling notation. Model-driven techniques have been used in the area of web development to a great extent. Most of the existing approaches are tuned toward specific platforms or develop only certain parts of web applications. These approaches generally use meta-models adapted to their targeted platforms. In order to flexibly target multiple platforms, the level of abstraction of the meta-model must be raised. Such a meta-model must allow the description of relevant features of web applications independently from the specificities of specific platforms. Additionally, transformations mapping from abstract to specific web descriptions must be expressible in a flexible way. In this thesis, we propose such an abstract meta-model. Mappings that transform abstract models to specific platforms are also presented. Different benefits can be foreseen from this approach. By relieving developers from low-level platform-specific related design, the approach has the potential to shift the development task to issues related to business needs. Another benefit is shortened development time. This could help web developers to overcome the problem of schedule delays, which is recognized as one of the top five most-cited problems with large-scale web systems. The approach is specifically suitable for information-intensive web-based systems. These applications typically involve large data stores accessed through a web interface. A distinctive aspect of this approach is its use of a specification of the data mapping as part of its high-level input. More importantly, the common features required to process data and communicate data objects between different layers and components are targeted.

In this thesis, we present an abstract meta-model for model driven development of web applications targeting multiple platforms. We review the existing technologies and the related work in order to obtain a list of requirements for such an abstract model. The abstract model is built by extending an existing UML-based model for web applications. We demonstrate that it is possible to map this abstract model to more than one specific development platform by providing transformations for these mappings. We also lay out the general outline of a model-driven process based on the proposed abstract model. The abstract model and the model-driven process are supported by a set of tools, case studies and a visual modeling notation. Model-driven techniques have been used in the area of web development to a great extent. Most of the existing approaches are tuned toward specific platforms or develop only certain parts of web applications. These approaches generally use meta-models adapted to their targeted platforms. In order to flexibly target multiple platforms, the level of abstraction of the meta-model must be raised. Such a meta-model must allow the description of relevant features of web applications independently from the specificities of specific platforms. Additionally, transformations mapping from abstract to specific web descriptions must be expressible in a flexible way. In this thesis, we propose such an abstract meta-model. Mappings that transform abstract models to specific platforms are also presented. Different benefits can be foreseen from this approach. By relieving developers from low-level platform-specific related design, the approach has the potential to shift the development task to issues related to business needs. Another benefit is shortened development time. This could help web developers to overcome the problem of schedule delays, which is recognized as one of the top five most-cited problems with large-scale web systems. The approach is specifically suitable for information-intensive web-based systems. These applications typically involve large data stores accessed through a web interface. A distinctive aspect of this approach is its use of a specification of the data mapping as part of its high-level input. More importantly, the common features required to process data and communicate data objects between different layers and components are targeted.

Made available in DSpace on 2014-06-12T15:51:08Z (GMT). No. of bitstreams: 2 arquivo2011_1.pdf: 1679074 bytes, checksum: c82d3e4381d7c70bcfaffe3bf1eddd78 (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2008 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Como qualquer outra tarefa evolucionária, a aplicação de refatoramentos em software orientado a objetos normalmente afeta código-fonte e seus modelos relacionados, aumentando a dificuldade de manutenção de artefatos corretos e consistentes. Devido à distância de representação entre artefatos de modelagem e programação, o esforço ligado a refatoramentos logo torna-se duplicado e custoso. Neste contexto, suporte de ferramentas utilizado atualmente, em especial ferramentas de Round-Trip Engineering (RTE), falha em automatizar tarefas de evolução. Consequentemente, a maioria dos projetos de software descarta artefatos de modelagem precocemente, adotando abordagens centradas unicamente em código-fonte. Esta tese propõe uma abordagem formal para consistentemente refatorar modelos de objeto e programas orientados a objetos, baseando o refatoramento apenas em modelos de objetos. Refatoramento de modelos é fundamentado com transformações formais primitivas { leis de modelagem { que são garantidamente preservadoras de semântica. Cada refatoramento aplicado a um modelo de objetos é associado a uma sequência semi-automática de aplicações de leis de programação preservadoras de comportamento, chamadas estrategias. Estrategias são aplicadas na dependência de um relacionamento especifico de conformidade entre modelos de objetos e programas, que devem satisfazer também um dado grau de confinamento. Este trabalho formaliza 14 estratregias, duas para cada lei de modelagem que afeta estruturas do programa. Estas estratregias são formalizadas como táticas de refinamento. Desta forma, refatoramento correto de programas pode ser realizado com reduzida intervenção manual do desenvolvedor, baseado apenas nas transformações que o mesmo aplicou ao modelo. Neste cenario, refatoramentos complexos que afetam as principais estruturas do programa podem ser aplicados a um artefato de mais alto nível de abstra ção, deixando a atualização semi-automática dos detalhes de implementação para as estratregias. Além disso, invariantes do modelo podem ser usados para aprimorar ferramentas especializadas em refatoramento, já que modelos de objetos oferecem informação semântica que permite refatoramentos automáticos mais poderosos. Esta tese considera Alloy como linguagem de modelagem formal, além de uma linguagem de programação similar a Java que chamamos BN. Para esta linguagem, introduzimos quatro novos refatoramentos e leis de programação orientada a objetos, com suas provas e derivações correspondentes. Adicionalmente, as leis de programação foram aplicadas em uma semântica de referências, mais próxima de linguages de programação utilizadas na prática. Com o intuito de delimitar a aplicabilidade desta abordagem, formalizamos uma noção de conformidade entre modelos de objetos e programas, a partir de um framework formal para definição de relacionamentos de conformidade; as definições formais relacionadas foram especificadas e checadas quanto ao tipo na ferramenta PVS. Além disso, estabelecemos e provamos manualmente um teorema para a corretude das estratregias, definindo que elas preservam comportamento e conformidade dos programas refatorados. Mesmo sendo uma abordagem formal, temos a preocupação de discutir sua utilização prática, além de aplica-la em três estudos de caso. Os problemas apresentados nesta tese certamente serão enfrentados em qualquer abordagem de desenvolvimento dirigida por modelos, no momento em que se lida com evolução

In this thesis, we present an abstract meta-model for model driven development of web applications targeting multiple platforms. We review the existing technologies and the related work in order to obtain a list of requirements for such an abstract model. The abstract model is built by extending an existing UML-based model for web applications. We demonstrate that it is possible to map this abstract model to more than one specific development platform by providing transformations for these mappings. We also lay out the general outline of a model-driven process based on the proposed abstract model. The abstract model and the model-driven process are supported by a set of tools, case studies and a visual modeling notation. Model-driven techniques have been used in the area of web development to a great extent. Most of the existing approaches are tuned toward specific platforms or develop only certain parts of web applications. These approaches generally use meta-models adapted to their targeted platforms. In order to flexibly target multiple platforms, the level of abstraction of the meta-model must be raised. Such a meta-model must allow the description of relevant features of web applications independently from the specificities of specific platforms. Additionally, transformations mapping from abstract to specific web descriptions must be expressible in a flexible way. In this thesis, we propose such an abstract meta-model. Mappings that transform abstract models to specific platforms are also presented. Different benefits can be foreseen from this approach. By relieving developers from low-level platform-specific related design, the approach has the potential to shift the development task to issues related to business needs. Another benefit is shortened development time. This could help web developers to overcome the problem of schedule delays, which is recognized as one of the top five most-cited problems with large-scale web systems. The approach is specifically suitable for information-intensive web-based systems. These applications typically involve large data stores accessed through a web interface. A distinctive aspect of this approach is its use of a specification of the data mapping as part of its high-level input. More importantly, the common features required to process data and communicate data objects between different layers and components are targeted.

To make computational thinking appealing to young learners, initial programming instruction looks very different now than a decade ago, with the increasing use of graphics and robots both real and virtual. After the first steps, children want to create interactive programs, and they need a model for this. State diagrams provide such a model, as observed previously by other researchers. This thesis documents the design and implementation of a Model-Driven Engineering tool, SD~Draw, that allows even primary-aged children to draw and understand state diagrams, and create modifiable app templates. We have tested this with grade 4 and 5 students. In our initial test, we discovered that children very quickly understand the motivation and use of state diagrams using this tool, and will independently discover abstract states even if they are only taught to model using concrete states. To determine whether this approach is appropriate for children of this age we asked three questions: do children understand state diagrams, do they understand the role of reachability, and are they engaged by them. We found that they are able to translate between different representations of the state diagrams, strongly indicating that they do understand them. We found with confidence p=0.001 that they do understand reachability by refuting the null hypothesis that they are creating diagrams randomly. And we found that they were engaged by the concept, with many students continuing to develop their diagrams on their own time after school and on the weekend. / Thesis / Master of Science (MSc)