No description available.

Action Theories are versatile and well-studied knowledge representation formalisms for modelling dynamic domains. However, traditional action theories allow only the specification of definite world knowledge, that is, universal rules for which there are no exceptions. When modelling a complex domain for which no complete knowledge can be obtained, axiomatisers face an unpleasant choice: either they cautiously restrict themselves to the available definite knowledge and live with a limited usefulness of the axiomatisation, or they bravely model some general, defeasible rules as definite knowledge and risk inconsistency in the case of an exception for such a rule. This thesis presents a framework for default reasoning in action theories that overcomes these problems and offers useful default assumptions while retaining a correct treatment of default violations. The framework allows to extend action theories with defeasible statements that express how the domain usually behaves. Normality of the world is then assumed by default and can be used to conclude what holds in the domain under normal circumstances. In the case of an exception, the default assumption is retracted, whereby consistency of the domain axiomatisation is preserved.

FROTA, Débora Farias. Uma lógica de descrição Default. 2011. 79 f. Dissertação (Mestrado em ciência da computação)- Universidade Federal do Ceará, Fortaleza-CE, 2011. / Submitted by Elineudson Ribeiro (elineudsonr@gmail.com) on 2016-07-11T13:21:13Z No. of bitstreams: 1 2011_dis_dffrota.pdf: 945021 bytes, checksum: 9adb958d87b14104dcd8db9fc4c4bd6f (MD5) / Approved for entry into archive by Rocilda Sales (rocilda@ufc.br) on 2016-07-15T13:25:42Z (GMT) No. of bitstreams: 1 2011_dis_dffrota.pdf: 945021 bytes, checksum: 9adb958d87b14104dcd8db9fc4c4bd6f (MD5) / Made available in DSpace on 2016-07-15T13:25:42Z (GMT). No. of bitstreams: 1 2011_dis_dffrota.pdf: 945021 bytes, checksum: 9adb958d87b14104dcd8db9fc4c4bd6f (MD5) Previous issue date: 2011 / Knowledge formalization and reasoning automatization are central within Arti cial Intelligence. First Order Logic has been traditionally used for such purposes. However, it is better suited to deal with complete knowledge in ideal circumstances. In real situations, in which the knowledge is partial, First Order Logic is not su cient. Nonmonotonic logics have been proposed to better cope with practical reasoning. A successful formalization of nonmonotonic reasoning is the Reiter's default logic which extends classical logic with default rules. Unfortunately, default logic is undecidable. In this work, we propose a description default logic expressible enough to formalize practical reasoning in knowledge bases. It has as its monotonic basis the ALC Description Logic. We add some restrictions to the application of defaults in order to obtain nice properties such as coherence and the elimination of anomalous extensions. We present the main algorithms used to build an extension with a step by step complexity analysis. / A formalização do conhecimento e a automatização do raciocínio são assuntos centrais de pesquisa da Inteligência Artificial. A Lógica de Primeira Ordem tem sido tradicionalmente utilizada para tais propósitos. No entanto, ela é mais adequada para lidar com conhecimento completo em circunstâncias ideais. Em situações reais, nas quais o conhecimento é parcial, a Lógica de Primeira Ordem não é suficiente. Lógicas não-monotônicas têm sido propostas para melhor lidar com o raciocínio prático. Uma formalização do raciocínio não-monotônico bem-sucedida é a Lógica Default de Reiter que estende a Lógica de Primeira Ordem com regras default. Infelizmente, a Lógica Default é indecidível. Nesta dissertação, propomos uma Lógica de Descrição Default expressiva o suficiente para formalizar o raciocínio prático sobre bases de conhecimento. Ela tem como base monotônica a Lógica de Descrição ALC. Adicionamos algumas restrições à aplicação dos defaults a fim de obter propriedades interessantes, tais como a coerência e a eliminação de extensões anômalas. Apresentamos os principais algoritmos usados para construir uma extensão com um passo-a-passo e suas análise de complexidade.

VERAS, Rodrigo de Melo Souza. Uma nova abordagem para o raciocinio não monotônico. 2007. 102 f. Dissertação (Mestrado em ciência da computação)- Universidade Federal do Ceará, Fortaleza-CE, 2007. / Submitted by Elineudson Ribeiro (elineudsonr@gmail.com) on 2016-07-12T16:47:33Z No. of bitstreams: 1 2007_dis_rmsveras.pdf: 1095797 bytes, checksum: 0e461d552478b8849c4691529b7a5a91 (MD5) / Approved for entry into archive by Rocilda Sales (rocilda@ufc.br) on 2016-07-21T16:12:24Z (GMT) No. of bitstreams: 1 2007_dis_rmsveras.pdf: 1095797 bytes, checksum: 0e461d552478b8849c4691529b7a5a91 (MD5) / Made available in DSpace on 2016-07-21T16:12:24Z (GMT). No. of bitstreams: 1 2007_dis_rmsveras.pdf: 1095797 bytes, checksum: 0e461d552478b8849c4691529b7a5a91 (MD5) Previous issue date: 2007 / Default logic was introduced to manipulate reasoning with incomplete information and became the main paradigm to formalize nonmonotonic reasoning. Many variations have been proposed with the objective to solve some limitations of the formalism or to consider different intuitions on the role of inconclusive information. However, some of the main characteristics had been kept: inconclusive information is represented as default rules, objectifies the calculation of extensions and, because this, they use characterizations through fixed-point operators. We consider a new approach for nonmonotonic reasoning. In this dissertation, we present the Defeasible Logic with Exception-First. The main advantages of this approach are: it does not use fixed-points operators to define expansions (our correspondent of extensions) and exception-first property that does not allow that a inconclusive proposition intervenes with the derivation of its exception. Moreover, we present a new way to define the extensions of default logic and two of its main variants, justified default logic and constrained default logic. / A lógica default foi introduzida para manipular raciocínio com conhecimento incompleto e tornou-se o principal paradigma para a formalização do raciocínio não monotônico. Muitas variações foram propostas com o objetivo de solucionar algumas limitações do formalismo ou para proporem diferentes intuições sobre o papel das informações inconlusivas. Porém, algumas das principais características foram mantidas: a informação inconclusiva é representada por regras default, objetivam o cálculo de extensões e, para isso, utilizam uma caracterização através de operadores de ponto fixo. Nós propomos uma nova abordagem para o raciocínio não monotônico. Nesta dissertação, apresentamos a Lógica Defeasible com Prioridade às Exceções. As principais vantagens desta proposta são a não utilização de pontos fixos para definição das expansões (nosso correspondente de extensões) e a propriedade de prioridade às exceções que não permite que uma proposição inconclusiva interfira na derivação de sua exceção. Além disso, apresentamos uma nova maneira de definir as extensões da lógica default e de duas das suas principais variantes a lógica default justificada e a lógica default com restrições.

nÃo hà / A lÃgica default foi introduzida para manipular raciocÃnio com conhecimento incompleto e tornou-se o principal paradigma para a formalizaÃÃo do raciocÃnio nÃo monotÃnico. Muitas variaÃÃes foram propostas com o objetivo de solucionar algumas limitaÃÃes do formalismo ou para proporem diferentes intuiÃÃes sobre o papel das informaÃÃes inconlusivas. PorÃm, algumas das principais caracterÃsticas foram mantidas: a informaÃÃo inconclusiva à representada por regras default, objetivam o cÃlculo de extensÃes e, para isso, utilizam uma caracterizaÃÃo atravÃs de operadores de ponto fixo. NÃs propomos uma nova abordagem para o raciocÃnio nÃo monotÃnico. Nesta dissertaÃÃo, apresentamos a LÃgica Defeasible com Prioridade Ãs ExceÃÃes. As principais vantagens desta proposta sÃo a nÃo utilizaÃÃo de pontos fixos para definiÃÃo das expansÃes (nosso correspondente de extensÃes) e a propriedade de prioridade Ãs exceÃÃes que nÃo permite que uma proposiÃÃo inconclusiva interfira na derivaÃÃo de sua exceÃÃo. AlÃm disso, apresentamos uma nova maneira de definir as extensÃes da lÃgica default e de duas das suas principais variantes a lÃgica default justificada e a lÃgica default com restriÃÃes. / Default logic was introduced to manipulate reasoning with incomplete information and became the main paradigm to formalize nonmonotonic reasoning. Many variations have been proposed with the objective to solve some limitations of the formalism or to consider different intuitions on the role of inconclusive information. However, some of the main characteristics had been kept: inconclusive information is represented as default rules, objectifies the calculation of extensions and, because this, they use characterizations through fixed-point operators. We consider a new approach for nonmonotonic reasoning. In this dissertation, we present the Defeasible Logic with Exception-First. The main advantages of this approach are: it does not use fixed-points operators to define expansions (our correspondent of extensions) and exception-first property that does not allow that a inconclusive proposition intervenes with the derivation of its exception. Moreover, we present a new way to define the extensions of default logic and two of its main variants, justified default logic and constrained default logic.

CoordenaÃÃo de AperfeiÃoamento de Pessoal de NÃvel Superior / A formalizaÃÃo do conhecimento e a automatizaÃÃo do raciocÃnio sÃo assuntos centrais de pesquisa da InteligÃncia Articial. A LÃgica de Primeira Ordem tem sido tradicionalmente utilizada para tais propÃsitos. No entanto, ela à mais adequada para lidar com conhecimento completo em circunstÃncias ideais. Em situaÃÃes reais, nas quais o conhecimento à parcial, a LÃgica de Primeira Ordem nÃo à suciente. LÃgicas nÃo-monotÃnicas tÃm sido propostas para melhor lidar com o raciocÃnio prÃtico. Uma formalizaÃÃo do raciocÃnio nÃo-monotÃnico bem-sucedida à a LÃgica Default de Reiter que estende a LÃgica de Primeira Ordem com regras default. Infelizmente, a LÃgica Default à indecidÃvel. Nesta dissertaÃÃo, propomos uma LÃgica de DescriÃÃo Default expressiva o suciente para formalizar o raciocÃnio prÃtico sobre bases de conhecimento. Ela tem como base monotÃnica a LÃgica de DescriÃÃo ALC. Adicionamos algumas restriÃÃes à aplicaÃÃo dos defaults a m de obter propriedades interessantes, tais como a coerÃncia e a eliminaÃÃo de extensÃes anÃmalas. Apresentamos os principais algoritmos usados para construir uma extensÃo com um passo-a-passo e suas anÃlise de complexidade. / Knowledge formalization and reasoning automatization are central within Articial Intelligence. First Order Logic has been traditionally used for such purposes. However, it is better suited to deal with complete knowledge in ideal circumstances. In real situations, in which the knowledge is partial, First Order Logic is not sucient. Nonmonotonic logics have been proposed to better cope with practical reasoning. A successful formalization of nonmonotonic reasoning is the Reiter's default logic which extends classical logic with default rules. Unfortunately, default logic is undecidable. In this work, we propose a description default logic expressible enough to formalize practical reasoning in knowledge bases. It has as its monotonic basis the ALC Description Logic. We add some restrictions to the application of defaults in order to obtain nice properties such as coherence and the elimination of anomalous extensions. We present the main algorithms used to build an extension with a step by step complexity analysis.

Action Theories are versatile and well-studied knowledge representation formalisms for modelling dynamic domains. However, traditional action theories allow only the specification of definite world knowledge, that is, universal rules for which there are no exceptions. When modelling a complex domain for which no complete knowledge can be obtained, axiomatisers face an unpleasant choice: either they cautiously restrict themselves to the available definite knowledge and live with a limited usefulness of the axiomatisation, or they bravely model some general, defeasible rules as definite knowledge and risk inconsistency in the case of an exception for such a rule. This thesis presents a framework for default reasoning in action theories that overcomes these problems and offers useful default assumptions while retaining a correct treatment of default violations. The framework allows to extend action theories with defeasible statements that express how the domain usually behaves. Normality of the world is then assumed by default and can be used to conclude what holds in the domain under normal circumstances. In the case of an exception, the default assumption is retracted, whereby consistency of the domain axiomatisation is preserved.

In this dissertation a formalisation of nonmonotonic reasoning, namely Default logic, is discussed. A proof theory for default logic and a variant of Default logic - Prioritised Default logic - is presented. We also pursue an investigation into the relationship between default reasoning and making inferences in a neural network. The inference problem shifts from the logical problem in Default logic to the optimisation problem in neural networks, in which maximum consistency is aimed at The inference is realised as an adaptation process that identifies and resolves conflicts between existing knowledge about the relevant world and external information. Knowledge and data are transformed into constraint equations and the nodes in the network represent propositions and constraint equations. The violation of constraints is formulated in terms of an energy function. The Hopfield network is shown to be suitable for modelling optimisation problems and default reasoning. / Computer Science / M.Sc. (Computer Science)

In this dissertation a formalisation of nonmonotonic reasoning, namely Default logic, is discussed. A proof theory for default logic and a variant of Default logic - Prioritised Default logic - is presented. We also pursue an investigation into the relationship between default reasoning and making inferences in a neural network. The inference problem shifts from the logical problem in Default logic to the optimisation problem in neural networks, in which maximum consistency is aimed at The inference is realised as an adaptation process that identifies and resolves conflicts between existing knowledge about the relevant world and external information. Knowledge and data are transformed into constraint equations and the nodes in the network represent propositions and constraint equations. The violation of constraints is formulated in terms of an energy function. The Hopfield network is shown to be suitable for modelling optimisation problems and default reasoning. / Computer Science / M.Sc. (Computer Science)

Cette thèse porte sur la modélisation du comportement humain dans les simulations de combat naval. Au sein de l'entreprise DCNS, les simulations de combat naval permettent d'évaluer les performances opérationnelles des navires militaires, dans un scénario donné. Les simulations actuelles ne permettent pas de prendre en compte l'analyse et la décision d'un opérateur, qui peuvent parfois conduire à des réactions inattendues. Le but de cette thèse est donc de modéliser le comportement d'un opérateur pour les simulations de combats navals.Pour représenter les connaissances, la logique non monotone la plus employée a été utilisée: la logique des défauts. Une prise en compte du temps a été ajoutée à cette logique des défauts. La logique des défauts va permettre de calculer des extensions. Chaque extension correspond à une action possible pour l'opérateur.Une méthode qui permet de choisir une extension a été définie. Cette méthode simule la décision de l'opérateur et elle prend en compte le caractère de l'opérateur. / This thesis deals with the modelling of operator behavior in naval action simulations. At DCNS, simulations of naval action estimate the operational performance of warships or submarines for a given scenario. However, the current simulations do not take into account the analysis and the decision of an operator, which can produce unexpected reactions. The purpose of this thesis is to develop a system allowing to model the behavior of an operatorin the naval action simulations.To represent knowledge, the most widely known non-monotonic logic is used : the default logic. A consideration of time is added to this logic. The default logic allows to calculate extensions. Each extension represents a possible action for the operator.A method to choose an extension is dened, which allows to simulate the decision of the operator and it handles the operator character.