In this thesis, we investigate the complexity of the membership problem for pattern languages. A pattern is a string over the union of the alphabets A and X, where X := {x_1, x_2, x_3, ...} is a countable set of variables and A is a finite alphabet containing terminals (e.g., A := {a, b, c, d}). Every pattern, e.g., p := x_1 x_2 a b x_2 b x_1 c x_2, describes a pattern language, i.e., the set of all words that can be obtained by uniformly substituting the variables in the pattern by arbitrary strings over A. Hence, u := cacaaabaabcaccaa is a word of the pattern language of p, since substituting cac for x_1 and aa for x_2 yields u. On the other hand, there is no way to obtain the word u' := bbbababbacaaba by substituting the occurrences of x_1 and x_2 in p by words over A. The problem to decide for a given pattern q and a given word w whether or not w is in the pattern language of q is called the membership problem for pattern languages. Consequently, (p, u) is a positive instance and (p, u') is a negative instance of the membership problem for pattern languages. For the unrestricted case, i.e., for arbitrary patterns and words, the membership problem is NP-complete. In this thesis, we identify classes of patterns for which the membership problem can be solved efficiently. Our first main result in this regard is that the variable distance, i.e., the maximum number of different variables that separate two consecutive occurrences of the same variable, substantially contributes to the complexity of the membership problem for pattern languages. More precisely, for every class of patterns with a bounded variable distance the membership problem can be solved efficiently. The second main result is that the same holds for every class of patterns with a bounded scope coincidence degree, where the scope coincidence degree is the maximum number of intervals that cover a common position in the pattern, where each interval is given by the leftmost and rightmost occurrence of a variable in the pattern. The proof of our first main result is based on automata theory. More precisely, we introduce a new automata model that is used as an algorithmic framework in order to show that the membership problem for pattern languages can be solved in time that is exponential only in the variable distance of the corresponding pattern. We then take a closer look at this automata model and subject it to a sound theoretical analysis. The second main result is obtained in a completely different way. We encode patterns and words as relational structures and we then reduce the membership problem for pattern languages to the homomorphism problem of relational structures, which allows us to exploit the concept of the treewidth. This approach turns out be successful, and we show that it has potential to identify further classes of patterns with a polynomial time membership problem. Furthermore, we take a closer look at two aspects of pattern languages that are indirectly related to the membership problem. Firstly, we investigate the phenomenon that patterns can describe regular or context-free languages in an unexpected way, which implies that their membership problem can be solved efficiently. In this regard, we present several sufficient conditions and necessary conditions for the regularity and context-freeness of pattern languages. Secondly, we compare pattern languages with languages given by so-called extended regular expressions with backreferences (REGEX). The membership problem for REGEX languages is very important in practice and since REGEX are similar to pattern languages, it might be possible to improve algorithms for the membership problem for REGEX languages by investigating their relationship to patterns. In this regard, we investigate how patterns can be extended in order to describe large classes of REGEX languages.

Padrões, linguagens de padrões e frameworks são formas de reuso de software. A complexidade de frameworks, causada pela dificuldade tanto em construí-los quanto em utilizá-los, é um dos inibidores do uso dessa tecnologia. Um processo para facilitar a construção e instanciação de frameworks é apresentado. Uma linguagem de padrões é utilizada para apoiar todo o processo, desde a identificação da funcionalidade do framework, seu projeto, implementação, validação, até sua instanciação para sistemas específicos do domínio. O framework obtido é do tipo caixa-branca e sua instanciação é feita especializando-se suas classes para sistemas específicos. O processo inclui, também, a construção de uma ferramenta para automatizar a instanciação do framework, por meio da qual é possível obter um sistema específico fornecendo apenas informações sobre os padrões da linguagem utilizados na sua modelagem. O processo é ilustrado com a Linguagem de Padrões para Gestão de Recursos de Negócios (GRN), que serviu de base para construção do framework GREN. Apresenta-se também o GREN-Wizard, uma ferramenta para instanciação automática do GREN para sistemas no domínio da GRN. A avaliação do processo é feita por meio de alguns experimentos e vários relatos de uso dos diversos sub-processos que compõem o processo geral, usando o GREN, a GRN e o GREN-Wizard. / Patterns, pattern languages, and frameworks are ways of enhancing software reuse. The complexity of software frameworks, caused both by the difficulty to build and to use them, is one of the inhibitors of this technology. A process to ease the construction and instantiation of frameworks is presented, in which a pattern language is used to support the whole process, starting from the identification of the framework functionality, its design, implementation, validation, and its instantiation to domain-specific systems. The framework that results from applying the proposed process is white-box and its instantiation is done by specializing its abstract classes according to a specific system. The process includes the construction of a tool to automate the framework instantiation, through which it is possible to obtain a specific system only by supplying information about the language patterns used to model it. The proposed process is illustrated with the Pattern Language for Business Resource Management (GRN), which was used as the basis for the GREN framework construction. The GREN-Wizard is also presented, which is a tool to automatically instantiate the GREN framework to specific applications in the GRN domain. The proposed process is evaluated by a few experiments and several uses of the sub-processes that compose the general process, using GRN, GREN and the GREN-Wizard.

Padrões, linguagens de padrões e frameworks são formas de reuso de software. A complexidade de frameworks, causada pela dificuldade tanto em construí-los quanto em utilizá-los, é um dos inibidores do uso dessa tecnologia. Um processo para facilitar a construção e instanciação de frameworks é apresentado. Uma linguagem de padrões é utilizada para apoiar todo o processo, desde a identificação da funcionalidade do framework, seu projeto, implementação, validação, até sua instanciação para sistemas específicos do domínio. O framework obtido é do tipo caixa-branca e sua instanciação é feita especializando-se suas classes para sistemas específicos. O processo inclui, também, a construção de uma ferramenta para automatizar a instanciação do framework, por meio da qual é possível obter um sistema específico fornecendo apenas informações sobre os padrões da linguagem utilizados na sua modelagem. O processo é ilustrado com a Linguagem de Padrões para Gestão de Recursos de Negócios (GRN), que serviu de base para construção do framework GREN. Apresenta-se também o GREN-Wizard, uma ferramenta para instanciação automática do GREN para sistemas no domínio da GRN. A avaliação do processo é feita por meio de alguns experimentos e vários relatos de uso dos diversos sub-processos que compõem o processo geral, usando o GREN, a GRN e o GREN-Wizard. / Patterns, pattern languages, and frameworks are ways of enhancing software reuse. The complexity of software frameworks, caused both by the difficulty to build and to use them, is one of the inhibitors of this technology. A process to ease the construction and instantiation of frameworks is presented, in which a pattern language is used to support the whole process, starting from the identification of the framework functionality, its design, implementation, validation, and its instantiation to domain-specific systems. The framework that results from applying the proposed process is white-box and its instantiation is done by specializing its abstract classes according to a specific system. The process includes the construction of a tool to automate the framework instantiation, through which it is possible to obtain a specific system only by supplying information about the language patterns used to model it. The proposed process is illustrated with the Pattern Language for Business Resource Management (GRN), which was used as the basis for the GREN framework construction. The GREN-Wizard is also presented, which is a tool to automatically instantiate the GREN framework to specific applications in the GRN domain. The proposed process is evaluated by a few experiments and several uses of the sub-processes that compose the general process, using GRN, GREN and the GREN-Wizard.

Knowledge Management Systems are deployed in organisations of all sizes to support the coordination and control of a range of intellectual assets, and the low cost infrastructures made available by the shift to ‘cloud computing’ looks to only increase the speed and pervasiveness of this move. However, their implementation has not been without its problems, and the development of novel interventions capable of supporting the mundane work of everyday organisational settings has ultimately been limited. A common source of trouble for those formulating such systems is said to be that some proportion of the knowledge held by a setting’s members is hidden from the undirected view of both The Organisation and its analysts - typically characterised as a tacit knowledge - and can therefore go unnoticed during the design and deployment of new technologies. Notwithstanding its utility, overuse of this characterisation has resulted in the inappropriate labelling of a disparate assortment of phenomena, some of which might be more appropriately re-specified as ‘hidden knowledges’: a standpoint which seeks to acknowledge their unspoken character without making any unwarranted claims regarding their cognitive status. Approaches which focus on the situated and contingent properties of the actual work carried out by a setting’s members - such as ethnomethodologically informed ethnography - have shown significant promise as a mechanism for transforming the role played by members’ practices into an explicit topic of study. Specifically they have proven particularly adept at noticing those aspects of members’ work that might ordinarily be hidden from an undirected view, such as the methodic procedures through which we can sometimes mean more than we can say in-just-so-many-words. Here - within the context of gathering the requirements for new Knowledge Management Systems to support the reuse of existing knowledge - the findings from the application of just such an approach are presented in the form of a Pattern Language for Knowledge Management Systems: a descriptive device that lends itself to articulating the role that such hidden knowledges are playing in everyday work settings. By combining these three facets, this work shows that it is possible to take a more meaningful approach towards noticing those knowledges which might ordinarily be hidden from view, and apply our new understanding of them to the design of Knowledge Management Systems that actively engage with the knowledgeable work of a setting’s members.

Massive Open Online Courses (MOOCs) hold the potential to open up educational opportunities and learning experiences to a global audience by combining recent technological advances with technology-enhanced learning. In general, MOOCs are considered online courses that require no prior qualifications for entry, can be accessed by anyone, and attract a diverse audience from a variety of learning and professional backgrounds. However, MOOC teams (including instructors and learning designers, among others) face several challenges when designing for learning in this context. In this work, two main challenges are investigated and approached, namely the lack of well-defined and validated learning design strategies to support practitioners in the MOOC development, and the poor pedagogical design models adopted in MOOCs, which are generally based on traditional classroom formats, such as teacher-centered approaches and content-based learning. Thus, the purpose of this work is to propose a learning design strategy, named Learning Design Framework for MOOCs (LDF4MOOCs), which is grounded on Software Engineering mechanisms and systematic procedures to ensure the standardization and the productivity of all the aspects involved in the MOOC development process. LDF4MOOCs consists of: (i) a MOOC Life Cycle process, which describes fundamental steps to plan, offer, and evaluate a MOOC; (ii) an Educational Design Pattern Language for MOOCs, which is based on problems and recurring solutions to solve the main activities described in the life cycle; and (iii) the related supporting resources. LDF4MOOCs is also pedagogically informed by Flipped Learning ideas, including active learning strategies, self-regulated learning, competency-based design, learner-centered learning, among others. LDF4MOOCs and its elements were evaluated through an experimental study, three case studies, and two expert reviews as internal evaluation methods. Additionally, a field evaluation with educators using the framework as a guide to design their MOOCs was considered as an external evaluation method. The obtained results indicated that LDF4MOOCs has a positive impact on the design for learning in MOOCs, suggesting that our strategy can be effectively applied to support and enhance MOOC development. / Cursos Online Abertos e Massivos (MOOCs) possuem o potencial de abrir oportunidades educacionais e experiências de aprendizado para um público global, combinando os avanços tecnológicos recentes e a aprendizagem mediada pela tecnologia. Em geral, eles são considerados cursos virtuais que não exigem qualificações prévias para a entrada, podem ser acessados por qualquer pessoa e atraem um público diversificado, com uma variedade de experiências e qualificações profissionais. No entanto, equipes responsáveis por desenvolverem MOOCs (incluindo instrutores e projetistas de aprendizagem, entre outros) deparam-se com vários desafios ao projetar para a aprendizagem nesse contexto. Neste trabalho, duas lacunas principais são investigadas e abordadas: a falta de estratégias de projeto de aprendizagem bem definidas e validadas para apoiar os profissionais no desenvolvimento de MOOCs; e as limitações nos modelos de projeto pedagógico adotados, geralmente baseados em formatos tradicionais de sala de aula, tais como abordagens centradas no professor e a aprendizagem baseada em conteúdo. O objetivo deste trabalho é propor e validar uma estratégia de projeto de aprendizagem denominada Learning Design Framework for MOOCs (LDF4MOOCs), baseada em mecanismos de Engenharia de Software e procedimentos sistemáticos para garantir a padronização e a produtividade de todos os aspectos envolvidos no processo de desenvolvimento de MOOCs. LDF4MOOCs consiste em: (i) um processo do ciclo de vida para MOOCs, que descreve etapas fundamentais para planejar, oferecer e avaliar um MOOC; (ii) uma Linguagem de Padrões de Projeto Educacional para MOOCs, baseada em problemas e soluções recorrentes para resolver as principais atividades descritas no ciclo de vida; e (iii) recursos de apoio relacionados. LDF4MOOCs também é pedagogicamente informado pelas ideias de Flipped Learning, incluindo estratégias de aprendizagem ativa, aprendizado autorregulado, projeto baseado em competências, aprendizado centrado no aluno, entre outros. O framework e seus elementos foram validados internamente por meio de um estudo experimental, três estudos de caso e duas revisões por especialistas. Adicionalmente, um estudo de campo envolvendo educadores que usaram o LDF4MOOCs como uma guia para desenvolver seus MOOCs foi utilizado como método de validação externa. Os resultados obtidos indicam que LDF4MOOCs apresenta um impacto positivo no projeto de aprendizagem para MOOCs, sugerindo que tal estratégia pode ser efetivamente aplicada para apoiar e melhorar o desenvolvimento de MOOCs.

Massive Open Online Courses (MOOCs) hold the potential to open up educational opportunities and learning experiences to a global audience by combining recent technological advances with technology-enhanced learning. In general, MOOCs are considered online courses that require no prior qualifications for entry, can be accessed by anyone, and attract a diverse audience from a variety of learning and professional backgrounds. However, MOOC teams (including instructors and learning designers, among others) face several challenges when designing for learning in this context. In this work, two main challenges are investigated and approached, namely the lack of well-defined and validated learning design strategies to support practitioners in the MOOC development, and the poor pedagogical design models adopted in MOOCs, which are generally based on traditional classroom formats, such as teacher-centered approaches and content-based learning. Thus, the purpose of this work is to propose a learning design strategy, named Learning Design Framework for MOOCs (LDF4MOOCs), which is grounded on Software Engineering mechanisms and systematic procedures to ensure the standardization and the productivity of all the aspects involved in the MOOC development process. LDF4MOOCs consists of: (i) a MOOC Life Cycle process, which describes fundamental steps to plan, offer, and evaluate a MOOC; (ii) an Educational Design Pattern Language for MOOCs, which is based on problems and recurring solutions to solve the main activities described in the life cycle; and (iii) the related supporting resources. LDF4MOOCs is also pedagogically informed by Flipped Learning ideas, including active learning strategies, self-regulated learning, competency-based design, learner-centered learning, among others. LDF4MOOCs and its elements were evaluated through an experimental study, three case studies, and two expert reviews as internal evaluation methods. Additionally, a field evaluation with educators using the framework as a guide to design their MOOCs was considered as an external evaluation method. The obtained results indicated that LDF4MOOCs has a positive impact on the design for learning in MOOCs, suggesting that our strategy can be effectively applied to support and enhance MOOC development. / Cursos Online Abertos e Massivos (MOOCs) possuem o potencial de abrir oportunidades educacionais e experiências de aprendizado para um público global, combinando os avanços tecnológicos recentes e a aprendizagem mediada pela tecnologia. Em geral, eles são considerados cursos virtuais que não exigem qualificações prévias para a entrada, podem ser acessados por qualquer pessoa e atraem um público diversificado, com uma variedade de experiências e qualificações profissionais. No entanto, equipes responsáveis por desenvolverem MOOCs (incluindo instrutores e projetistas de aprendizagem, entre outros) deparam-se com vários desafios ao projetar para a aprendizagem nesse contexto. Neste trabalho, duas lacunas principais são investigadas e abordadas: a falta de estratégias de projeto de aprendizagem bem definidas e validadas para apoiar os profissionais no desenvolvimento de MOOCs; e as limitações nos modelos de projeto pedagógico adotados, geralmente baseados em formatos tradicionais de sala de aula, tais como abordagens centradas no professor e a aprendizagem baseada em conteúdo. O objetivo deste trabalho é propor e validar uma estratégia de projeto de aprendizagem denominada Learning Design Framework for MOOCs (LDF4MOOCs), baseada em mecanismos de Engenharia de Software e procedimentos sistemáticos para garantir a padronização e a produtividade de todos os aspectos envolvidos no processo de desenvolvimento de MOOCs. LDF4MOOCs consiste em: (i) um processo do ciclo de vida para MOOCs, que descreve etapas fundamentais para planejar, oferecer e avaliar um MOOC; (ii) uma Linguagem de Padrões de Projeto Educacional para MOOCs, baseada em problemas e soluções recorrentes para resolver as principais atividades descritas no ciclo de vida; e (iii) recursos de apoio relacionados. LDF4MOOCs também é pedagogicamente informado pelas ideias de Flipped Learning, incluindo estratégias de aprendizagem ativa, aprendizado autorregulado, projeto baseado em competências, aprendizado centrado no aluno, entre outros. O framework e seus elementos foram validados internamente por meio de um estudo experimental, três estudos de caso e duas revisões por especialistas. Adicionalmente, um estudo de campo envolvendo educadores que usaram o LDF4MOOCs como uma guia para desenvolver seus MOOCs foi utilizado como método de validação externa. Os resultados obtidos indicam que LDF4MOOCs apresenta um impacto positivo no projeto de aprendizagem para MOOCs, sugerindo que tal estratégia pode ser efetivamente aplicada para apoiar e melhorar o desenvolvimento de MOOCs.

Doctor of Philosophy / The thesis of this study is that two seemingly disparate research disciplines can be coalesced to develop an effective pedagogical framework for educational design in the context of networked learning. That contention is grounded in, and inspired by, the rapid developments in educational technologies which have greatly changed the landscape in teaching and learning in higher education over the last decade. The study attempts to add to the corpus of contemporary learning theory which sees students not merely as passive recipients of knowledge, but as active participants in the learning process, having much greater control over their selection of technological learning tools, learning resources and learning methodologies. This is very much in line with the shift from the traditional focus on content design and knowledge transmission towards a more student-centred design for knowledge co-construction, a development which demands the type of new thinking about the design of learning tasks and learning resources contained in this study. Also set out are new lines of action for the fashioning of a collaborative learning environment, for community interaction and the sharing of knowledge, and for promoting good teaching and learning practice. The central argument of the study is that such pedagogical goals may be attained by juxtaposing the theories of Systemic Functional Linguistics (hereafter SFL) and pattern languages. These have not, thus far, been used in combination. SFL is a well established theory in the study of language, and is used in this thesis to help analyse and classify discourses produced and shared by teachers and students in networked learning. Pattern languages have their origin in architecture. Design patterns can be used as a means of representing and sharing important and specific empirical research results and design experiences. This new knowledge can be used to support and improve the quality of educational design. The study has two central components. The first uses the SFL theoretical framework to demonstrate how text is used as a key medium in networked learning. In other words, it is argued in this section that the quality of texts has a direct impact on the quality of learning and learning outcomes. The quality of text is assessed by means of a detailed discourse analysis of selected texts. This process involves deconstructing, identifying and capturing the linguistic resources and language strategies used in the texts. The detailed discourse analysis also illustrates and reveals how language is used in the construction of knowledge and the promotion of collaboration in teaching and learning. The second component centres on the argument that SFL provides valuable language knowledge which can be represented by using Alexander’s design patterns. New knowledge encoded in these design patterns can be used by teachers and designers as reusable and shared resources to help them improve their design work. The empirical research was carried out in three phases. The first involved a) the identification of text patterns of discourses used in networked learning based on detailed discourse analysis; b) Interviewing experienced academic staff to identify their perspectives on good online teaching practices and success factors. The second phase involved using the data which emerged from these interviews and discourse analysis to model illustrative patterns. (Here, illustrative means that due to the scope of the study, it is only possible to develop a limited number of patterns to illustrate the methods used for pattern development. It is not the intention to develop a full repository of design patterns in this study). In the third (validation) phase the patterns were reviewed by two groups of academic staff, with the aim of improving these patterns. Improved patterns were then tested on a group of educational design students for their usefulness and application. It is concluded from this research that it is possible to develop design patterns which ensure the best use of linguistic resources in both the teaching and learning process. Finally, it is argued that the combination of SFL and pattern languages provides a promising theoretical framework for the complex and demanding task of educational design. Future research could make use of such a framework to explore a fuller application of the pattern- based approach for the representation of new knowledge for educational design. Suggested additional research directions include finding new ways of capturing a new pedagogical approach to mobile learning and blended learning. Also, a promising direction could be the use of SFL Appraisal theory (Martin, 2000) for the investigation on how students construct interpersonal relationships (appraise peer work) in online joint projects. In the conclusion, it is contended that through its exploration of new ground in the use of SFL and pattern language theory in the construction of education design patterns, the study makes a significant contribution to knowledge in the field of networked learning.

Doctor of Philosophy / The thesis of this study is that two seemingly disparate research disciplines can be coalesced to develop an effective pedagogical framework for educational design in the context of networked learning. That contention is grounded in, and inspired by, the rapid developments in educational technologies which have greatly changed the landscape in teaching and learning in higher education over the last decade. The study attempts to add to the corpus of contemporary learning theory which sees students not merely as passive recipients of knowledge, but as active participants in the learning process, having much greater control over their selection of technological learning tools, learning resources and learning methodologies. This is very much in line with the shift from the traditional focus on content design and knowledge transmission towards a more student-centred design for knowledge co-construction, a development which demands the type of new thinking about the design of learning tasks and learning resources contained in this study. Also set out are new lines of action for the fashioning of a collaborative learning environment, for community interaction and the sharing of knowledge, and for promoting good teaching and learning practice. The central argument of the study is that such pedagogical goals may be attained by juxtaposing the theories of Systemic Functional Linguistics (hereafter SFL) and pattern languages. These have not, thus far, been used in combination. SFL is a well established theory in the study of language, and is used in this thesis to help analyse and classify discourses produced and shared by teachers and students in networked learning. Pattern languages have their origin in architecture. Design patterns can be used as a means of representing and sharing important and specific empirical research results and design experiences. This new knowledge can be used to support and improve the quality of educational design. The study has two central components. The first uses the SFL theoretical framework to demonstrate how text is used as a key medium in networked learning. In other words, it is argued in this section that the quality of texts has a direct impact on the quality of learning and learning outcomes. The quality of text is assessed by means of a detailed discourse analysis of selected texts. This process involves deconstructing, identifying and capturing the linguistic resources and language strategies used in the texts. The detailed discourse analysis also illustrates and reveals how language is used in the construction of knowledge and the promotion of collaboration in teaching and learning. The second component centres on the argument that SFL provides valuable language knowledge which can be represented by using Alexander’s design patterns. New knowledge encoded in these design patterns can be used by teachers and designers as reusable and shared resources to help them improve their design work. The empirical research was carried out in three phases. The first involved a) the identification of text patterns of discourses used in networked learning based on detailed discourse analysis; b) Interviewing experienced academic staff to identify their perspectives on good online teaching practices and success factors. The second phase involved using the data which emerged from these interviews and discourse analysis to model illustrative patterns. (Here, illustrative means that due to the scope of the study, it is only possible to develop a limited number of patterns to illustrate the methods used for pattern development. It is not the intention to develop a full repository of design patterns in this study). In the third (validation) phase the patterns were reviewed by two groups of academic staff, with the aim of improving these patterns. Improved patterns were then tested on a group of educational design students for their usefulness and application. It is concluded from this research that it is possible to develop design patterns which ensure the best use of linguistic resources in both the teaching and learning process. Finally, it is argued that the combination of SFL and pattern languages provides a promising theoretical framework for the complex and demanding task of educational design. Future research could make use of such a framework to explore a fuller application of the pattern- based approach for the representation of new knowledge for educational design. Suggested additional research directions include finding new ways of capturing a new pedagogical approach to mobile learning and blended learning. Also, a promising direction could be the use of SFL Appraisal theory (Martin, 2000) for the investigation on how students construct interpersonal relationships (appraise peer work) in online joint projects. In the conclusion, it is contended that through its exploration of new ground in the use of SFL and pattern language theory in the construction of education design patterns, the study makes a significant contribution to knowledge in the field of networked learning.

Made available in DSpace on 2016-06-02T19:06:01Z (GMT). No. of bitstreams: 1 4801.pdf: 4117819 bytes, checksum: b4df67024157ee1a2c79256315a97e7d (MD5) Previous issue date: 2012-11-13 / Universidade Federal de Sao Carlos / Software development techniques continually evolve in order to improve development and maintenance processes in addition to lower costs and higher quality. The goal of MDD is to reduce the semantic distance between a problem and its solution specification. Therefore MDD focuses on high-level abstraction modeling and successive model transformations, until finally, generate code. Studies assert that model-driven development can be significantly more efficient than traditional source code-driven software development and still reduce the possibility of occurrence of several problems during the software life-cycle. Likewise, Web engineering can also be benefited by MDD adoption, especially when supported by approaches that facilitate MDD use. Web development is usually agile with frequent releases, these approaches must be flexible to adapt to this context. However, generally, the approaches proposed by the academic community have complex processes which involve many different model definitions, programming languages, plug-ins and IDEs. These features contradict the practices adopted by Web developers. This paper presents the MWDA (Model-Driven Web Applications) approach that provides a simple process to support model-driven web development. This approach does not depend on tools, technologies or plug-ins and encourage combination with other forms of reuse and development processes. Furthermore, the Ruby- MDWA was developed with Ruby language and Ruby on Rails framework support, in order to create Web applications with MDWA assistance. This tool provides a set of textual models and defines M2M and M2C transformation tools, maintaining the requirements traceability since its specification to its construction and further maintenance. In order to show the use of the approach and tool, it was performed a real study case with a software company, from São Carlos SP, where a project management system was developed. In parallel, two experiments were conducted with undergraduate students in Computer Science and Computer Engineering and a Masters in Computer Science, to evaluate the gains and limitations of the Ruby-MDWA tool. / As técnicas de desenvolvimento de software evoluem continuamente com a finalidade de melhorar processos de construção e manutenção de software, além de obter ganhos em tempo, custo e qualidade. O objetivo do MDD é reduzir a distância semântica entre um problema e a especificação de sua solução. Para isso, MDD tem enfoque na modelagem de alto nível de abstração e em sucessivos refinamentos dos modelos construídos em artefatos mais detalhados, até enfim, gerar código. Há afirmações de que o desenvolvimento orientado a modelos pode ser significativamente mais eficiente que o desenvolvimento tradicional guiado por código fonte, além de reduzir a possibilidade de ocorrência de uma série de problemas durante o ciclo de vida do software. Da mesma forma, a engenharia de aplicações Web também pode ser beneficiada pela adoção de MDD, em especial com o apoio de abordagens que facilitem sua utilização. Como o desenvolvimento de aplicações Web comumente é ágil e com publicações freqüentes, essas abordagens devem ser flexíveis para que se adaptem a esse contexto. Entretanto, em geral, as abordagens propostas pela comunidade acadêmica apresentam processos complexos que envolvem diversos modelos, linguagens de programação, plug-ins e ambientes de programação. Essas características contrariam as práticas aprovadas pelos desenvolvedores Web. Esta dissertação apresenta a abordagem MDWA (Model-Driven Web Applications) que fornece um processo simples para desenvolvimento de software Web com apoio de MDD. A abordagem não depende de ferramentas, tecnologias ou plug-ins e estimula a combinação com outras formas de reuso e processos de desenvolvimento. Além disso, foi construída uma ferramenta, denominada Ruby-MDWA, baseada na linguagem Ruby e no framework Ruby on Rails destinada à criação de aplicações Web com auxílio da abordagem MDWA. Essa ferramenta fornece um conjunto de quatro modelos textuais e define transformadores M2M e M2C, que mantém a rastreabilidade de um requisito desde sua especificação até sua construção e posterior manutenção. Para mostrar o uso da abordagem e da ferramenta, foi realizado um estudo de caso real em conjunto com uma empresa de software de São Carlos SP, onde um sistema de gerenciamento de projetos foi desenvolvido. De forma paralela, foram conduzidos dois experimentos com alunos de graduação em Bacharelado em Ciência da Computação e Engenharia de Computação e mestrado em computação da UFSCar, visando avaliar os ganhos e as limitações da ferramenta Ruby-MDWA.