L’objectif de ce travail a consisté à évaluer différents effets écotoxiques, cytotoxique et le potentiel androgénique de 14 produits pharmaceutiques: la clomipramine, la fluoxétine, la fluvoxamine, la paroxétine, la sertraline et la venlafaxine (antidépresseurs), l’amphotéricine B, l’éconazole, le kétoconazole et le miconazole (antifongiques), la clarithromycine (antibiotique), l’acébutolol (ß-bloquants), le gabapentine (antiépileptiques) et la cétirizine (antihistaminiques), à l’aide d’études in vivo et in vitro. Les tests d’écotoxicité aiguë (96h) sur le cnidaire d’eau douce Hydra attenuata ont montré un effet délétère sur la survie des hydres adultes, la morphologie et le développement des embryons pour l’ensemble d’antidépresseurs et des antifongiques testés. Les valeurs de CE50 calculées ont révélé que l’éconazole, le miconazole et la sertraline sont les molécules les plus toxiques. Les tests d’écotoxicité chronique sur les hydres adultes (10 jours) et les embryons (14 jours) ont mis en évidence que la plus part des molécules d’antidépresseurs testées sont des embryotoxiques et tératogènes puisque ils induisent des malformations chez les embryons contrairement aux antifongiques qui présentent des anomalies seulement chez les adultes. Ces tests ont montré aussi que les molécules d’antidépresseurs testées seules sont moins toxiques que lorsqu’elles sont testées en mélange. En effet, le mélange de molécules antidépresseurs testées présente une toxicité plus élevée sur les adultes que sur les embryons, avec une concentration minimale induisant des effets significatifs de l’ordre de ng L-1, concentrations proches de celles retrouvées dans l’environnement. De plus, les résultats d’écotoxicité chronique sur plusieurs générations d’hydres ont montré que la morphologie des individus de la première génération (F0) ainsi que ceux de la deuxième génération (F1) été affectée de la même manière, démontrant une sensibilité du même ordre vis-à-vis de la sertraline ou du mélange d’antidépresseurs, tandis que les individus issus de la troisième génération (F2) apparaissent moins sensibles lors de l'exposition à la sertraline. Par contre, la reproduction des hydres exposées à la sertraline est diminué de plus en plus dans chaque génération, avec une toxicité plus importante pour les individus de la troisième génération. Le mélange d’antidépresseurs affecte la reproduction des hydres de la deuxième génération F1 à des concentrations environnementales. Les essais in vitro conduits sur les mêmes molécules pharmaceutiques, testées seules et en mélange, sur les cellules MDA-Kb2 n'ont révélé aucun effet androgénique et cytotoxique sur les cellules à des concentrations environnementales / The objective of this work was to evaluate different ecotoxic, cytotoxic and androgenic effects of 14 pharmaceutical compounds: clomipramine, fluoxetine, fluvoxamine, paroxetine, sertraline and venlafaxine, amphotericin B, éconazole, ketoconazole and miconazole, clarithromycin, the acebutolol, gabapentin and cetirizine, using in vivo and in vitro studies. The acute ecotoxicity tests (96h) on the freshwater cnidarian Hydra attenuata showed deleterious effects on adult survival, morphology and embryo development for all tested antidepressants and antifungals drug. The values of EC50 concentrations revealed that econazole, miconazole and sertraline are the most toxic molecules. The chronic ecotoxicity tests on hydras adults (10 days) and embryos (14 days) revealed that most of the molecules tested antidepressants are embryotoxic and teratogenic as they cause malformations in embryos contrary to antifungal which display abnormalities in adults only. These tests also showed that antidepressants tested individually are less toxic than when tested as a mixture. This mixture has a higher toxicity on adults than on embryos, with significant effects concentrations around ng L-1 which are close to measured environmental concentrations. In addition, chronic ecotoxicity results over several generations of hydras have shown identical sensitivity of both first and second generation on morphology of individuals. No more toxicity was observed on the third generation of hydra. Concerning the reproduction of hydras, this parameters was strongly decreased in all generations with an increasing sensitivity of individuals along the generations. Moreover, the mixture of antidepressants affect reproduction of hydras for the second generation at environmental concentrations. In vitro tests conducted on MDA-KB2 cells with the same pharmaceuticals revealed that these compounds tested did not exert any androgenic and cytotoxic effect on cells in culture at environmental concentrations

Made available in DSpace on 2015-04-22T22:00:54Z (GMT). No. of bitstreams: 1 Dissertacao_Alyson.pdf: 2045006 bytes, checksum: 552fa4e2c7d4a7aa118296a3b4324aa2 (MD5) Previous issue date: 2011-09-09 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The Wireless Sensor Networks (WSNs) revolutionize the monitoring of environment, enabling the creation of applications and scenarios in various areas of knowledge. As a consequence, we have a very fast growth of the development of new applications and rata increase of system complexity and cost of proposed development. The difficulty in developing new applications is not only due to restrictive features (memory, processor, power), but also by the instructions of low-level programming these networks. This situation is conductive to the creation of new methodologies and tools that support the development of system for that plataform. This work proposes the development of a tool called automatic code generation for wireless sensor network (WSN-GAC), which has as its main objective to enable any product high in developing applications for WSNs using models specified in Model driven architecture (MDA). WSN-GAC receives as input a model of a component-based application created by the designer and then performs automatic code generation of nesC (network embedded system C). Three case studied demonstrate the potential of WSN-GAC to help developers of applications for these networks. / As Redes de Sensores Sem Fio (RSSFs) revolucionam a monitoração de ambientes, possibilitando a criação de aplicações e cenários nas mais diversas áreas do conhecimento. Como consequência, temos o crescimento muito rápido do desenvolvimento de novas aplicações e proporcionalmente aumento de complexidade dos sistemas propostos e do custo de desenvolvimento. A dificuldade no desenvolvimento de novas aplicações não se deve somente às características restritivas (memória, processamento, energia), mas também pelas instruções de programação de baixo nível destas redes. Este panorama é favorável a criação de novas metodologias e ferramentas que dêem suporte ao desenvolvimento de sistemas para tal plataforma. Este trabalho propõe o desenvolvimento de uma ferramenta intitulada Geração Automática de Código para Redes de Sensores Sem Fio (GAC-RSSFs), que tem como objetivo principal possibilitar a alta produtividade no desenvolvimento de aplicações para as RSSFs utilizando modelos especificados na Arquitetura Orientada a Modelos (Model Driven Architecture - MDA). GAC-RSSFs recebe como entrada um modelo baseado em componentes de uma aplicação criada pelo projetista e então realiza a geração automática de código em linguagem nesC (network embedded systems C). Três estudos de caso demonstram o potencial de GAC-RSSFs para ajudar os desenvolvedores de aplicações para essas redes.

Made available in DSpace on 2016-08-17T14:52:51Z (GMT). No. of bitstreams: 1 Samyr Beliche Vale.pdf: 1271783 bytes, checksum: 3556a7f5490acfc5fdf7431663fba9fc (MD5) Previous issue date: 2004-03-10 / This paper presents a methodology for integrating business applications with Web services through the creation of meta-models in MDA. This methodology covers steps for the integration process and presents an architecture based on semantic middlware to integrate legacy applications. / Este trabalho apresenta uma metodologia para integração de aplicações empresariais com web services através da criação de meta-modelos em MDA. Esta metodologia abrange fases para o processo de integração e apresenta uma arquitetura baseada em middlware semântico para integrar aplicações legadas.

Made available in DSpace on 2016-08-17T14:52:52Z (GMT). No. of bitstreams: 1 GEYSA HELENA GUIMARAES CHAVES.pdf: 3886332 bytes, checksum: 1e032419dfdc24d8a137ce9f3e0c992e (MD5) Previous issue date: 2008-06-26 / This research work presents a proposal for constructing and maintenance of trusting to be perceived by a pretense consumer in a process of making decisions and by means of the involved risks in commercializing in a unsafe environment, the Internet. The dissertation examine the concept of trust as well as the factors and processes that contribute to it in order to elicitate aspects that the consumer consider as relevant to deposit their trusting in an online seller. The aspects that were captured and represented by a metamodel of trusting allows the creation of independent platform models. Thus, the model of trusting can be reused and easily modified for future developments and integrations. An editor of model in the form of trees trusting is provided as plug-in for Eclipse. The mechanisms described as builders of trusting are evidences, information in the form of metrics that enable the reduction of uncertainties in an on-line transactions. Finally, a scenario that allows the description of the results obtained in this research work is presented. / Este trabalho apresenta uma proposta para construção e manutenção de confiança a ser percebida por um pretenso consumidor em um processo de tomada de decisão e mediante aos riscos envolvidos em comercializar em um ambiente inseguro, a Internet. Esta dissertação examina a natureza do conceito de confiança, bem como os fatores e processos que contribuem para a mesma, a fim de elicitar elementos que o consumidor considera como relevantes ao depositar sua crença em um vendedor on-line. Esses elementos foram capturados e representados através de um metamodelo de confiança que permite a criação de modelos independentes de plataformas. Assim, o modelo de confiança pode ser reutilizado e facilmente modificado para futuras evoluções e integrações. Um editor de modelos de confiança em forma de árvore é fornecido como plug-in para Eclipse. Os mecanismos descritos como construtores de confiança são evidências, informações em forma de métricas que possibilitam a diminuição do grau de incerteza em transações on-line. Por fim, um cenário que possibilita a descrição dos resultados obtidos na realização deste trabalho dissertativo é apresentado.

L'interopérabilité est une des caractéristiques requises pour les entreprises évoluant dans un marché globalisé à la concurrence croissante et complexe. Dans la dernière décennie, l'interopérabilité des entreprises a été développée et prescrite par différents types de cadres, de méthodes et de techniques. Cependant, le développement de l'interopérabilité n'est pas encore assez mature pour être considéré en tant que science à part entière. Par ailleurs, il ne cesse d'évoluer en fonction des besoins des entreprises,de leurs environnements et des différents secteurs d’activité. Aujourd'hui, l'environnement s’organise en réseaux multipleet provoque d’imprévisibles situations liées à leurs dynamiques (création, modification, résilience). Ainsi l’interopérabilité durable devient une dimension nouvelle de recherche pour l'interopérabilité des systèmes d'entreprise et de leurs domaines d'applications. Dans l'interopérabilité durable, l'interopérabilité d'entreprise dynamique est l'un des thèmes focaux. Cette approche dynamique, également appelée « fédérée », est originaire du cadre d'interopérabilité de l'Entreprise proposée dans le Réseau d’Excellence (NoE)INTEROP. Il vise à donner la capacité aux entreprises d’établir une interopérabilité à la volée sans connaissance préalable des informations à échanger. Cette thèse présente l'état actuel des travaux qui se rapprochent du développement de l'interopérabilité des entreprises « fédérés » en dynamique. Ces travaux de thèse mettent tout d’abord en évidence l’intérêt de la redécouverte de modèles à partir d’un système existant avant de concevoir un futur système. Une méthodologie de réverse engineering dirigée par les modèles et basée sur la norme de simulation distribuée HLA est proposée pour concevoir et développerpar l'approche fédérée d'interopérabilité le futur système d’information de l’entreprise. La phase de mise en œuvre réutilise les concepts d’interopérabilité issusde la simulation distribuée pour faciliter et coordonner la communication entre les systèmes d'information distribués hétérogènes des entreprises en combinant avec les dernières orientation service actuelle du web. La plate-forme tend ainsi à satisfaire les attentes de la dernière version du standard de l'architecture de haut niveau HLA 1516 Evolved. Ce cadre propose donc un cycle complet de développement pour qui a l'intention de réutiliser un système d'information existant sans recoder ex-nilo, mais en l’adaptant aux nouvelles exigences de la dynamique d'interopérabilité. / Interoperability is one of the requisite features for existing enterprises in the increasing competitive and complex global market. In the last decade, enterprise interoperability has been developed and prescribed by various kinds of frameworks, methods, and techniques. However interoperability development is still not mature enough to become a science. Meanwhile, it keeps evolving according to different business requirement and market environment. Nowadays, networked environment causes unpredictable dynamical situations, thus sustainable interoperability becomes a new research dimension in the interoperability of enterprise systems and applications domain. In the sustainable interoperability, enterprise interoperability dynamics is one of the focal topics. This dynamic approach also called federated is originated from Enterprise Interoperability Framework of INTEROP NoE, which aims to establish interoperability on the fly. This thesis presents current state on federated approaches to develop enterprise interoperability dynamics. Based on this study, a reversible model driven and HLA based methodology is proposed for achieving federated approach for Enterprise Interoperability. It reuses distributed simulation interoperability concepts to facilitate and coordinate the communication between heterogeneous distributed information systems of the enterprises. The platform is complaint with the latest version of the High Level Architecture (HLA) that is a distributed communication standard. This framework is also proposing a development lifecycle that intends to reuse existing information systems without recoding them but by adapting them to the new requirements of interoperability dynamics.

The Model Driven Architecture (MDA) allows moving the software development from the time consuming and error-prone level of writing program code to the next higher level of modeling. In order to gain benefit from this innovative technology, it is necessary to satisfy two requirements. These are first, the creation of compact, complete and correct platform independent models (PIM) and second, the development of a flexible and extensible model transformation framework taking into account frequent changes of the target platform. In this thesis a platform-based methodology is developed to create PIM by abstracting common modeling elements into a platform independent modeling library called Design Platform Model (DPM). The DPM contains OCL-based types for modeling primitive and collection types, a platform independent GUI toolkit as well as other common modeling elements, such as those for IO-operations. Furthermore, a DPM profile containing diverse domain specific and design pattern-based stereotypes is also developed to create PIM with high-level semantics. The behavior in PIM is specified using an OCL-like action language called eXecutable OCL (XOCL), which is also developed in this thesis. For model transformation, the model compiler MOCCA is developed based on a flexible and extensible architecture. The model mapper components in the current version of MOCCA are able to map desktop applications onto JSE platform; the both business object layer and persistence layer of a three-layered enterprise applications onto JEE platform and SAP ABAP platform. The entire model transformation process is finished with complete code generation.

The Model Driven Architecture (MDA) allows moving the software development from the time consuming and error-prone level of writing program code to the next higher level of modeling. In order to gain benefit from this innovative technology, it is necessary to satisfy two requirements. These are first, the creation of compact, complete and correct platform independent models (PIM) and second, the development of a flexible and extensible model transformation framework taking into account frequent changes of the target platform. In this thesis a platform-based methodology is developed to create PIM by abstracting common modeling elements into a platform independent modeling library called Design Platform Model (DPM). The DPM contains OCL-based types for modeling primitive and collection types, a platform independent GUI toolkit as well as other common modeling elements, such as those for IO-operations. Furthermore, a DPM profile containing diverse domain specific and design pattern-based stereotypes is also developed to create PIM with high-level semantics. The behavior in PIM is specified using an OCL-like action language called eXecutable OCL (XOCL), which is also developed in this thesis. For model transformation, the model compiler MOCCA is developed based on a flexible and extensible architecture. The model mapper components in the current version of MOCCA are able to map desktop applications onto JSE platform; the both business object layer and persistence layer of a three-layered enterprise applications onto JEE platform and SAP ABAP platform. The entire model transformation process is finished with complete code generation.

Initié en 2009, le projet MISE 2.0 (deuxième itération du projet Mediation Information System Engineering) s’articule autour d’une approche BPM (pour Business Process Management) et d’une vision MDE (pour Model-Driven Engineering). La réalisation d’une démarche BPM classique au sein d’une organisation nécessite de recueillir une connaissance couvrant à la fois les aspects structurel, informationnel et fonctionnel afin de définir des modèles de processus caractéristiques du comportement de l’organisation. Concernant le projet MISE 2.0, l’approche BPM considérée concerne un ensemble d’organisations collaboratives. Quant à la composante MDE, elle est destinée à faciliter l’automatisation des différentes étapes de la démarche : i) Recueil de la connaissance (caractérisation de la situation) : Il s’agit de collecter les information concernant la situation collaborative considérée, ii) Déduction de la cartographie de processus collaboratifs (définition de la solution) : il s’agit de définit les processus collaboratifs adaptés à la situation collaboratives caractérisée au niveau précedent and iii) Déploiement du SI de médiation (implémentation de la solution) : il s’agit d’implémenter le SI de médiation sous la forme d’une plateforme informatique capable d’orchestrer les processus collaboratif définis. La problématique scientifique relève des deux transitions entre ces trois niveaux d’abstractions : la première transition est prise en charge au niveau abstrait de la démarche MISE 2.0 alors que la seconde est traitée au niveau concret. Les travaux de thèse dont il est ici question se focalisent sur le niveau abstrait : déduction d’une cartographie de processus collaboratifs satisfaisant la situation collaborative considérée. Ce type d’objectif relève généralement d’activités entièrement manuelles qui nécessitent une importante quantité de travail afin d’obtenir les modèles de processus escomptés. Les travaux de recherches présentés ambitionnent d’automatiser cette démarche. Le principe est le suivant : (i) recueil, sous la forme de modèles, de la connaissance nécessaire à la caractérisation de la situation collaborative (informations sur les partenaires, les fonctions qu’ils partagent et leurs objectifs), (ii) déduction de la connaissance complémentaire relative à la dynamique collaborative qui pourrait satisfaire ces objectifs selon les moyens disponibles (cette phase s’appuie sur un métamodèle collaboratif, sur l’ontologie associée et sur des règles de transformation) et (iii) structuration de cette connaissance générée sous la forme d’une cartographie de processus collaboratifs (grâce à des algorithmes dédiés). / MISE 2.0 (for Mediation Information System Engineering, second iteration) project has been launched in 2009. The MISE 2.0 engineering approach is based on BPM (Business Process Management) and MDE (Model-Driven Engineering). Running a regular BPM approach on a specific organization consists in gathering structural, informational, and functional knowledge in order to design cartography of processes covering the behavior of the modeled organization. Regarding the MISE 2.0 project the BPM approach concerns a set of organizations and MDE helps in automatizing the different steps: i) Knowledge gathering (situation layer): collect information concerning the collaborative situation, ii) Processes cartography design (solution layer): design the processes according to the knowledge gathered and iii) MIS deployment (implementation layer): implement an IT structure able to run the processes cartography. Both the transitions between these layers are the hard-points of this approach: The first gap is managed at the abstract level of MISE 2.0 while the second one is managed at the concrete level of MISE 2.0. The current PhD is focused on the first issue: designing a relevant processes cartography from the modeled collaborative situation. However, this is usually a manual activity, which requires a large amount of work to draw the processes and their links. The current research works aim at building such collaborative process cartography in an automated manner. Our principles are (i) to gather the essential and minimum initial collaborative knowledge (e.g. partners, shared functions and collaborative objectives) in models, ii) to deduce the missing knowledge with the help of a collaborative metamodel, an associated ontology and transformation rules and iii) to structure the deduced knowledge in a collaborative process cartography thanks to dedicated algorithms.

[pt] Sistemas multi-agentes (SMA) diferem de sistemas que não são baseados em agentes devido ao fato dos agentes serem unidades autônomas capazes de ações flexíveis e inteligentes. Por este motivo é proposto na literatura um grande número de metodologias, frameworks e linguagens para dar apoio a construção deste tipo de sistema. Muitas dessas metodologias e suas ferramentas provêm da comunidade de inteligência artificial e estão focadas em uma arquitetura específica de agentes. Este trabalho propõe o uso da arquitetura Model Driven Architecture (MDA), descrita pela OMG, no processo de desenvolvimento de SMA. MDA é uma arquitetura para desenvolvimento de software estruturada em etapas de modelagem que dá suporte a todo o ciclo de desenvolvimento de um sistema. O processo de desenvolvimento proposto está dividido de acordo com as etapas apresentadas por MDA. Na etapa PIM, onde são especificados modelos independentes de plataforma, propomos a utilização da linguagem de modelagem MAS-ML para SMA. Na etapa PSM, onde são definidos modelos específicos de plataforma, propomos utilizar a linguagem de modelagem UML. Os modelos MAS-ML definidos na etapa PIM são transformados em modelos UML na etapa PSM com base em um framework para implementação de SMA utilizando orientação a objetos. Na ultima etapa do desenvolvimento, o código da aplicação é gerado a partir dos modelos UML. Este trabalho detalha as etapas PIM e PSM do processo de desenvolvimento de SMA e as transformações necessárias para a geração de código. Para demonstrar a aplicação do processo proposto são apresentadas duas aplicações de SMA desenvolvidas utilizando o mesmo. Ao final, apresenta-se a ferramenta MAS-ML Tool desenvolvida para dar apoio ao processo de desenvolvimento proposto. Esta ferramenta implementa todas as etapas presentes no processo, permitindo assim a implementação de SMA a partir de sua modelagem. / [en] Multi-agent systems (MAS) differ from non-agent systems because agents are intended to be autonomous units capable of flexible and intelligent actions. For this reason it is proposed in the literature a great number of methodologies frameworks and languages to support the development of these systems. Several methodologies and their tools are come from artificial intelligent community and are focused in a specific agent architecture. This work proposes the use of the Model Driven Architecture (MDA), described by OMG, in the development process of MAS. MDA specifies a structured software development process in modeling stages that supports all system development life cycle. The proposed development process is divided according to the MDA stages. In PIM stage, where platform independent models are specified, we propose the use of MAS-ML modeling language for MAS. In PSM stage, where platform specific models are specified, we propose the use of UML modeling language. The MASML models defined on PIM stage are transformed in UML models at PSM stage, based on an object-oriented framework for implementing MAS. In the last development stage, the application code is generated from UML models. This work details the PIM and PSM stages of the MAS development process and the models transformations to generate source code. To exemplify the applicability of the proposed MAS development process, two different MAS applications were developed based on the process. Finally, a MAS-ML tool is presented. Such tool was developed to support the proposed development process. The tool implements all stages presented in the process, allowing the modeling and implementation of MAS.

Este trabalho tem por objetivo propor a utilização de soluções pré-definidas e comprovadas na abordagem MDA, auxiliando o desenvolvedor na resolução de problemas recorrentes ao desenvolvimento de software. A utilização destas soluções pré-definidas leva a um aumento de produtividade no contexto MDA e na geração de software baseado em boas práticas. Para que este objetivo seja cumprido, é empreendida uma análise sobre o MDA e sobre como operacionalizar as transformações entre modelos. Também é empreendida uma análise sobre o suporte fornecido pela UML e sobre reutilização em desenvolvimento orientado a modelos. Por fim este trabalho apresenta partes de uma aplicação protótipo, construída para ser uma prova de conceito de código gerado através da combinação de UML e soluções pré-definidas. / The goal of this work is to propose the use of pre-defined solutions on MDA approach, supporting the developer in solving recurrent problems of software development. The use of these pre-defined solutions leads to an increase of productivity in MDA context, and in the generation of software based on best practices. To reach this goal, an analysis of MDA is undertaken, as well as an analysis of how to enable the transformations between models. It is also undertaken an analysis about the use of UML and the reuse in model driven development. Finally this work presents portions of a prototype application, constructed to be a proof-of-concept of generated code, combining UML and the pre-defined solutions.