Made available in DSpace on 2016-06-02T19:05:53Z (GMT). No. of bitstreams: 1 3986.pdf: 2744978 bytes, checksum: 19e350743f147492505973082c83902d (MD5) Previous issue date: 2009-08-28 / Financiadora de Estudos e Projetos / Aiming at improving competitive advantage, organizations in general have been giving more attention to knowledge engineering, as well as its techniques and methods to acquire and represent knowledge already in place. This knowledge is sometimes not fully explored, or not explored properly. This approach is to develop a knowledge modeling strategy for reactive production scheduling, which focuses on tacit and explicit knowledge acquisition and representation. The knowledge representation aims at getting results closer to the organization reality and should also be comprehensible and easy to be validated by experts. To validate the proposed model, a computational system was developed to evaluate the model behavior under a specific domain, as well as to verify if the results are satisfactory. The model supports decision-making when unplanned events occur during the production process, enabling the possibility of evaluating greater knowledge to make the decision and also allowing the operator to make the decision without any expert support. The validation was performed in two steps; the first one, the computational system validation, was carried out by unit tests on every module, and after that, an integration test was performed. The second one, the proposed model validation, was verified in two ways. The first one was the validation using the industry use cases based on industry platform existing in the TEAR laboratory; the acceptance and understanding of the models by experts were verified using the same models already validated in a project between an industry and the TEAR laboratory. / Buscando uma vantagem competitiva, as empresas têm dado maior atenção à engenharia do conhecimento, visando o uso de técnicas para extrair e representar os conhecimentos já existentes, porém algumas vezes inexplorados ou explorados de maneira inadequada. A proposta deste trabalho é apresentar uma estratégia de modelagem do conhecimento no âmbito da programação reativa da produção. Essa estratégia de modelagem visa realizar a aquisição tanto do conhecimento tácito quanto do explícito, para representá-los de uma maneira que obtenha resultados mais próximos da realidade, visando também um tipo de representação que seja de fácil compreensão para a validação dos especialistas do domínio. Com o objetivo de validar o modelo proposto, foi desenvolvido um sistema computacional, para avaliar o comportamento do sistema perante dados do domínio e verificar se os resultados obtidos por ele estão compatíveis com o desejado. Esse modelo visa auxiliar a tomada de decisão dos especialistas perante eventos não programados que possam ocorrer durante a execução da programação, possibilitando a tomada de decisão do especialista, de uma maneira mais abrangente, pelo fato de considerar o conhecimento de vários especialistas e também possibilitando a tomada de decisão pelo operador na falta do especialista da área. A validação desse trabalho se deu em dois momentos, o primeiro sendo a validação do sistema computacional que foi realizado módulo por módulo, e depois a integração deles. O segundo foi a validação do modelo proposto em dois aspectos, o primeiro validado por meio dos casos de uso levantados tendo como base a fábrica presente no laboratório TEAR, e o outro aspecto foi a facilidade do entendimento dos modelos pelos especialistas, seguindo os mesmos métodos de representação utilizados em um projeto existente entre o laboratório TEAR e uma indústria.

O grande avanço das técnicas de Planejamento em Inteligência Artificial fez com que a Engenharia de Requisitos e a Engenharia do Conhecimento ganhassem extrema importância entre as disciplinas relacionadas a projeto de engenharia (Engineering Design). A especificação, modelagem e análise dos domínios de planejamento automático se tornam etapas fundamentais para melhor entender e classificar os domínios de planejamento, servindo também de guia na busca de soluções. Neste trabalho, é apresentada uma proposta de um ambiente integrado de modelagem e análise de domínios de planejamento, que leva em consideração o ciclo de vida de projeto, representado por uma ferramenta gráfica de modelagem que utiliza diferentes representações: a UML para modelar e analisar as características estáticas dos domínios; XML para armazenar, integrar, e exportar informação para outras linguagens (ex.: PDDL); as Redes de Petri para fazer a análise dinâmica; e a PDDL para testes com planejadores. / The great development in Artificial Intelligence Planning has emphasized the role of Requirements Engineering and Knowledge Engineering among the disciplines that contributes to Engineering Design. The modeling and specification of automated planning domains turn out to be fundamental tasks in order to understand and classify planning domains and guide the application of problem solving techniques. In this work, it is presented the proposed integrated environment for modeling and analyzing automated planning domains, which considered the life cycle of a project, represented by a tool that uses several language representations: UML to model and perform static analyses of planning environments; XML to hold, integrate, share and export information to other language representations (e.g. PDDL); Petri Nets, where dynamic analyses are made; and PDDL for testing models with planners.

Made available in DSpace on 2016-08-17T14:53:06Z (GMT). No. of bitstreams: 1 UIRATAN ALVES DE SOUSA CAVALCANTE.pdf: 4868154 bytes, checksum: 71d6895bc849a94e0bac2dba2b713adf (MD5) Previous issue date: 2009-12-01 / The agent-oriented development paradigm has achieved a high maturity level over the last decade. However, integrated development environments supporting all phases of agent-oriented development are still missing. MADAE-Pro is a process for the development and reuse of family of multi-agent systems and integrates two complementary process. One is based on Domain Engineering concepts, aiming in creating artifacts of reusable software in the development of an application family in a particular domain problem, and the other is based on Application Engineering, which guides the construction of applications based on reusable software artifacts previously produced in the Domain Engineering process. The environment includes the ONTORMAS ontology and, through inference rules and semantic search over its instances, automates the modeling tasks of the MADAE-Pro process. These features allow for the developer productivity gains and for maintaining the consistence of the knowledge-base, ensuring the success of the reuse in future applications. / O paradigma de desenvolvimento orientado a agentes tem atingido um alto nível de maturidade na última década. Entretanto, ainda faltam ambientes integrados de desenvolvimento de software que suportem todas as fases do desenvolvimento orientado a agentes. MADAE-Pro é um processo para o desenvolvimento e reuso de famílias de sistemas multiagente e integra dois subprocessos complementares. Um é baseado nos conceitos da Engenharia de Domínio, isto é, visa construir artefatos reutilizáveis que representem uma família de aplicações e o outro, baseado na Engenharia de Aplicações, guia o desenvolvimento de uma aplicação específica reutilizando os produtos do primeiro subprocesso. Este trabalho propõe MADAEIDE, um ambiente integrado de desenvolvimento baseado no conhecimento que automatiza o processo desenvolvimento de software multiagente MADAE-Pro. O ambiente agrega a ontologia ONTORMAS e, através de regras de inferência e buscas semânticas diretamente sobre suas instâncias, automatiza as tarefas de modelagem do processo MADAE-Pro. Estas características levam a um ganho de produtividade por parte do desenvolvedor, além de manter uma base de conhecimento consistente, garantido o sucesso do reuso em futuras aplicações.

Ce travail de thèse, réalisé dans le cadre d’une convention CIFRE avec l’entreprise Essilor et le laboratoire ERPI de l’Université de Lorraine, propose un cadre d’évolution des organisations basée sur le management opérationnel des connaissances et des compétences. Nous avons mis en valeur les mécanismes organisationnels de partage et de mise en oeuvre des connaissances et compétences en environnement de conception, afin de proposer des modèles, méthodes et outils permettant aux acteurs et organisations d’améliorer leurs performances. Quatre aspects ont été développés dans ce travail, dont trois contributions et une évaluation de l’efficacité du cadre d’évolution proposé : • L’élaboration d’une méthode de modélisation d’entreprise basée sur le modèle conceptuel KROM et de sa déclinaison en modèles opérationnels. Cette méthode vise à expliciter les liens organisationnels des concepts de connaissance et de compétence afin de les rattacher à leurs contextes de mise en oeuvre. • La définition d’un cadre global d’évolution d’une organisation sur la base du management de ses connaissances et compétences. Ce cadre d’évolution comprend une caractérisation et une évaluation de la maturité des organisations dans leurs pratiques de gestion des connaissances et des compétences, ainsi qu’une méthode d’évolution des organisations reposant sur la proposition de scénarios d’actions d’amélioration : la méthode ACKME. • La proposition d’une approche organisationnelle de conception de Systèmes Multi-agents (SMA) appelée DOCK, permettant de spécifier des SMA adaptés aux enjeux rencontrés par les approches d’ingénierie à base de connaissances, à savoir la prise en compte et la gestion du cycle de vie de la connaissance. • Une évaluation de l’intérêt du cadre d’évolution mis en oeuvre (les trois contributions précédentes) sur la base de l’évaluation des compétences des acteurs de l’organisation. Cette évaluation repose d’une part sur la possibilité d’intégrer le cadre d’évolution dans une approche globale de pilotage du changement dans l’organisation et d’autre part, sur des expérimentations menées au sein du service Ingénierie d’Essilor. / This research, carried out under a PhD “CIFRE” with the company ESSILOR and the ERPI Laboratory of the Lorraine University, proposes a framework to make organizations evolve, based on the operational management of knowledge and competences. We have enhanced the organizational mechanisms of sharing, and the use of knowledge and competences within the design environment, in order to propose models, methods and tools allowing actors and organizations to improve their performances. Four key elements have been developed in this research, including three contributions and an assessment of the efficiency of the proposed development framework: • The development of an enterprise modeling method based on the conceptual model KROM and its translation into operational models. This method aims to clarify the organizational links of the knowledge and competence concepts in order to link them to their implementation contexts. • The definition of a global framework to make an organization evolve based on the management of its knowledge and competence assets. This evolution framework includes a characterization and an assessment of the organizations’ maturity in their knowledge and competences management practices, as well as a method to make organizations change, based on the proposal of scenario of improvement actions: the ACKME method. • The proposal of an organizational approach to design Multi-agent systems (MAS) called DOCK, allowing the designer to specify MAS oriented toward the issues faced by knowledge engineering approaches, namely the recognition and management of the knowledge life cycle. • An assessment of the value of the implemented framework (the three previous contributions) based on the assessment of the competences of the actors within the organization. This evaluation is based, on one hand, on the possibility to integrate the evolution framework into a global steering approach to make the organization evolve, and on the other hand, on experiments led within the Engineering department of Essilor

La thèse, qui pose la question de la description du roman et envisage une réponse par l’interrogation d’une problématique documentaire (proposition d’un modèle théorique de représentation, base pour la conception et la réalisation d’outils de navigation raisonnée dans la fiction) conduit une réflexion sur la place de la fiction littéraire dans la construction des savoirs et, par extension, d’une culture. La réflexion et l’expérimentation se font dans le contexte scolaire du lycée d’enseignement général (définition du statut du roman dans l’utilisation qui en est faite dans les programmes et objectifs d’enseignement, analyse de la notion de lecture littéraire en lycée) et par l’analyse d’un corpus de textes : étude de La mystérieuse flamme de la Reine Loana / Umberto Eco, 2004, traduction française, Paris, Grasset et Fasquelle, 2005, doublée de l’étude d’un corpus aléatoire sur le thème de la Shoah. L’objet d’étude étant le roman, les liens entre les Sciences de l’Information et de la Communication et la sémiotique sont abordés sous un jour nouveau : construction d’une théorie sémiotique (définition, mise en place d’un langage formel, choix d’un système de représentation) adaptée à l’objet d’étude par l’explicitation des conditions de production de sens, ce qui implique une hybridation des champs (SIC et théories d’analyse littéraire). Sont développés les éléments théoriques pour un modèle de représentation de la fiction. Les définitions du roman, de la fiction littéraire narrative, de l’information, de la fictiologie (nouveau champ de recherche ouvert par l’étude) et de l'information fictiologique critique, complétées par une description du positionnement dans le champ des Sciences de l’Information et de la Communication, posent le cadre de la réflexion. Un chapitre traitant de la littérature et de la notion de mémoire permet de décrire les liens entre roman et représentation de la réalité. Pour cela, sont définies et étudiées les notions de texte, de référence et de contexte. Enfin, est explicité le titre de la thèse par la description de la faculté imaginale et de la fonction de médiation prêtées au roman. Pour clôturer cette partie d’étude, un état de l’art de la classification de la fiction est opéré par l’élaboration d’un panorama des outils proposés et utilisés, renforcé par la définition d’une recherche ouverte d’information et une étude de la notion de sens dans le traitement de la fiction. Les analyses dans le champ pratique comportent un compte-rendu d’enquêtes, menées dans le cadre de l’enseignement secondaire et de la formation des documentalistes de l’Education Nationale, sur les pratiques de lecture. L’analyse des corpus de textes permet d’interroger la problématique de la réalité visible dans le roman. Ces travaux expérimentent une ontologie fictiologique par une analyse du roman comme document et un essai de formalisation sur la littérature des camps. La proposition d’un modèle théorique à expérimenter se concrétise par l’élaboration d’un système d’organisation de connaissances info-littéraires. Une réflexion sur l’environnement technique actuel justifie les choix retenus, les propositions et solutions réfléchies. En présentant une littératie fictiologique, prolongement de la littératie médiatique, la thèse se clôture par la description d’une médiation fictiologique qui s’ancre dans une renégociation des espaces et des typologies documentaires, investiguant les pratiques participatives et collaboratives. / The dissertation, which investigates the description of the novel and considers an answer by interrogating an information problem (offering a theoretical model of representation, a basis for the conception and creation of tools for a carefully thought out navigation within fiction), carries out a reflection about the role of literary fiction in the construction of bodies of knowledge and, by extension, of a culture. These reflection and experimentation take place in the school context of the “lycee d’enseignement general” [mainstream education high school] (definition of the status of the novel in the way that it is used in education programs and objectives, analysis of the notion of literary reading in high school) and through the analysis of a body of texts: study of La Mysterieuse Flamme de la Reine Loana / Umberto Eco, 2004, French translation, Paris, Grasset et Fasquelle, 2005, combined with the study of an unpredictable collection of texts about the theme of the Shoah. Since the object of study is the novel, the links between Information and Communication Sciences and semiotics are approached in a new light: construction of a semiotic theory (definition, setting up of a formal language, choice of a representation system) made suitable to the object of study by clarifying the conditions of meaning production, which implies a hybridization of the fields (Information and Communication Sciences and literary analysis). The theoretical elements for a representation model of fiction are developed. Definitions of the novel, of literary narrative fiction, of information, of fictiology (a new field of research opened up by the study) and of critical fictiological information, complemented by a description of positioning of this research in the field of Information and Communication Sciences, frame the reflection. A chapter tackling literature and the notion of memory allows for a delineation of the links between novel and representation of reality. For this purpose, the notions of text, reference, and context are defined. Finally, the title of the dissertation is explained through the description of the imaginal faculty, and of the mediation function attributed to the novel. To close this part of the study, drafting a panorama of the proposed and employed tools helps operate a state of the art of fiction classification, reinforced by the definition of an open information research and a study of the notion of meaning in the handling of fiction. Analyses in the practical field include a report of investigations on reading practices conducted in the contexts of secondary education and of the training of ‘Education Nationale’ information librarians. Analyzing bodies of texts allows the problem of visible reality in the novel to be interrogated. This research tests out a fictiologic ontology by examining the novel as document and by attempting a formalization of the camps literature. The proposition of a theoretical model to be experimented is concretized by the development of a system of organization of information and literary knowledge. A reflection on the current technical environment justifies the choices considered, the propositions and solutions offered. By presenting a fictiological information literaty, an extension of media information literaty, the dissertation ends with the description of fictiological mediation anchored into a renegotiation of information spaces and typologies, investigating participatory and collaborative practices.