A resolução de problemas é uma metodologia de investigação utilizada pela Psicologia experimental desde sua fundação. Um debate na área é quais os efeitos de variáveis de treino sobre a resolução de problemas, e como ocorre um tipo específico de resolução, a chamada resolução súbita, tradicionalmente descrita como Insight. O presente trabalho teve por objetivo averiguar a resolução de problemas a partir do treino independente de habilidades pré-requisito em diferentes tarefas de recombinação de repertórios com diferentes espécies. Adicionalmente, foram feitos testes de entendimento de propriedades funcionais de objetos e repertórios treinados. No Experimento I, dois pombos (Columba livia) foram testados no procedimento padrão de deslocamento de caixa. Os sujeitos aprenderam: a) empurrar uma caixa em direção a um alvo, e b) subir e bicar outro alvo. O teste final requeria que o animal empurrasse a caixa para debaixo de um alvo, subisse e o bicasse. Os dois sujeitos resolveram a tarefa final. No Experimento II, quatro pombos passaram por histórias assimétricas de treino das duas habilidades pré-requisito também da tarefa de deslocamento de caixa. Apenas um dos sujeitos resolveu o problema, e sem a mesma fluidez apresentada pelos sujeitos do Experimento I. O Experimento III manipulou a consequência utilizada no treino de cada habilidade pré-requisito, também na tarefa de deslocamento de caixas. Empurrar em direção a uma luz foi treinado com alimento como consequência, e subir e bicar foi treinado com água como consequência. Nenhum dos quatro pombos resolveu a tarefa em sua primeira tentativa. Todos os sujeitos que resolveram a tarefa nos Experimentos I, II ou III apresentaram resultados similares nos testes de entendimento de propriedades funcionais, que indicaram que pequenas mudanças na situação de teste dificultam a resolução. No Experimento IV, com ratos (Rattus norvegicus), uma nova tarefa foi desenvolvida, que envolvia a recombinação de a) cavar maravalha, e b) subir escadas. Na tarefa final os ratos deveriam cavar a maravalha, encontrar uma passagem escondida que dava acesso a um segundo ambiente no qual haviam escadas que levavam a um pedaço de alimento. Dois ratos aprenderam os dois repertórios e resolveram a tarefa, de forma súbita. Outros quatro ratos que aprenderam somente um dos repertórios não resolveram a tarefa. O Experimento V testou os efeitos da quantidade de treino (Experimento II), tipo de consequência (Experimento III) e contexto de treino e teste, em uma tarefa específica com corvos da Nova Caledônia (Corvus moneduloides). Todos os animais resolveram a tarefa final, entretanto, de forma não súbita. Os dados indicam que a resolução súbita, a partir da recombinação de repertórios aprendidos independentemente é passível do efeito de uma série de variáveis de treino e teste até então pouco estudadas, como a quantidade de treino, tipo de consequência e contexto de treino e teste / Experimental Psychology utilizes problem-solving methodology since its inception. Much was debated about the effects of training variables upon the solution of a problem, especially concerning the sudden solution of specific problem, called Insight. The present work aimed to investigate the solution of different problem in different species after the independent training of pre-requisite repertoires. Additionally, tests regarding the understanding of functional properties of stimuli were made. In Experiment I, two pigeons (Columba livia) were tested in the box-displacement test. The subjects learned to, a) push a box towards a target, and b) to climb and peck another target. The final task required to subjects to push one box towards a target hanging in the ceiling, climb the box and then peck the target. The two subjects solved the final task. In Experiment II, four pigeons had an asymmetric training of the two abilities. Only one of the subjects solved the final task, not as fluidly as the subjects from Experiment I. In Experiment III, the consequence used during the training of each ability was different. Pushing towards a target was reinforced with food, and to climb and peck with water. None of the four birds solved the task in its first presentation. All the subjects that solved the tasks in Experiments I, II and III showed similar performances in the tests regarding the understanding of functional properties of stimuli, suggesting that even minima changes in the test situation can impair the solution. Experiment IV employed a new task, with rats (Rattus norvegicus). Subjects learned to a) digg shavings, and b) climb stairs. In the final task, the rats had to digg and find a passage leading to another ambient containing stairs that lead to food. Two rats that learned the two abilities solved the problem, suddenly. Other four rats that learned only one of the two abilities (dig or climb) didnt solved the problem. Experiment V tested the effects of quantities of training (Experiment II), consequences of trained abilities (Experiment III), and context of training and testing with new caledonian crows (Corvus moneduloides). All four crows solved the problem, but not suddenly. The data of all experiments suggests that the sudden solution, when acquired through the interconnection of repertoires is more subtle then it is usualy regarded, and that variables of training and testing are crucial for the solution of a problem

The performance of Internet-based services depends on many server-side, client-side, and network related factors. Often, the interaction among the factors or their effect on service performance is not known or well-understood. The complexity of these services makes it difficult to develop analytical models. Lack of models impedes network management tasks, such as predicting performance while planning for changes to service infrastructure, or diagnosing causes of poor performance. We posit that we can use statistical causal methods to model performance for Internet-based services and facilitate performance related network management tasks. Internet-based services are well-suited for statistical learning because the inherent variability in many factors that affect performance allows us to collect comprehensive datasets that cover service performance under a wide variety of conditions. These conditional distributions represent the functions that govern service performance and dependencies that are inherent in the service infrastructure. These functions and dependencies are accurate and can be used in lieu of analytical models to reason about system performance, such as predicting performance of a service when changing some factors, finding causes of poor performance, or isolating contribution of individual factors in observed performance. We present three systems, What-if Scenario Evaluator (WISE), How to Improve Performance (HIP), and Network Access Neutrality Observatory (NANO), that use statistical causal methods to facilitate network management tasks. WISE predicts performance for what-if configurations and deployment questions for content distribution networks. For this, WISE learns the causal dependency structure among the latency-causing factors, and when one or more factors is changed, WISE estimates effect on other factors using the dependency structure. HIP extends WISE and uses the causal dependency structure to invert the performance function, find causes of poor performance, and help answers questions about how to improve performance or achieve performance goals. NANO uses causal inference to quantify the impact of discrimination policies of ISPs on service performance. NANO is the only tool to date for detecting destination-based discrimination techniques that ISPs may use. We have evaluated these tools by application to large-scale Internet-based services and by experiments on wide-area Internet. WISE is actively used at Google for predicting network-level and browser-level response time for Web search for new datacenter deployments. We have used HIP to find causes of high-latency Web search transactions in Google, and identified many cases where high-latency transactions can be significantly mitigated with simple infrastructure changes. We have evaluated NANO using experiments on wide-area Internet and also made the tool publicly available to recruit users and deploy NANO at a global scale.

No description available.

No description available.

Le concept de confiance est un concept sociocognitif qui adresse la question de l'interaction dans les systèmes concurrents. Quand la complexité d'un système informatique prohibe l'utilisation de solutions traditionnelles de sécurité informatique en amont du processus de développement (solutions dites de type dur), la confiance est un concept candidat, pour le développement de systèmes d'aide à l'interaction. Dans cette thèse, notre but majeur est de présenter une vue d'ensemble de la discipline de la modélisation de la confiance dans les systèmes informatiques, et de proposer quelques modèles logiques pour le développement de module de confiance. Nous adoptons comme contexte applicatif majeur, les applications basées sur les architectures orientées services, qui sont utilisées pour modéliser des systèmes ouverts telle que les applications web. Nous utiliserons pour cela une abstraction qui modélisera ce genre de systèmes comme des systèmes multi-agents. Notre travail est divisé en trois parties, la première propose une étude de la discipline, nous y présentons les pratiques utilisées par les chercheurs et les praticiens de la confiance pour modéliser et utiliser ce concept dans différents systèmes, cette analyse nous permet de définir un certain nombre de points critiques, que la discipline doit aborder pour se développer. La deuxième partie de notre travail présente notre premier modèle de confiance. Cette première solution basée sur un formalisme logique (logique dynamique épistémique), démarre d'une interprétation de la confiance comme une croyance sociocognitive, ce modèle présentera une première modélisation de la confiance. Apres avoir prouvé la décidabilité de notre formalisme. Nous proposons une méthodologie pour inférer la confiance en des actions complexes : à partir de notre confiance dans des actions atomiques, nous illustrons ensuite comment notre solution peut être mise en pratique dans un cas d'utilisation basée sur la combinaison de service dans les architectures orientées services. La dernière partie de notre travail consiste en un modèle de confiance, où cette notion sera perçue comme une spécialisation du raisonnement causal tel qu'implémenté dans le formalisme des règles de production. Après avoir adapté ce formalisme au cas épistémique, nous décrivons trois modèles basés sur l'idée d'associer la confiance au raisonnement non monotone. Ces trois modèles permettent respectivement d'étudier comment la confiance est générée, comment elle-même génère les croyances d'un agent et finalement, sa relation avec son contexte d'utilisation. / The concept of trust is a socio-cognitive concept that plays an important role in representing interactions within concurrent systems. When the complexity of a computational system and its unpredictability makes standard security solutions (commonly called hard security solutions) inapplicable, computational trust is one of the most useful concepts to design protocols of interaction. In this work, our main objective is to present a prospective survey of the field of study of computational trust. We will also present two trust models, based on logical formalisms, and show how they can be studied and used. While trying to stay general in our study, we use service-oriented architecture paradigm as a context of study when examples are needed. Our work is subdivided into three chapters. The first chapter presents a general view of the computational trust studies. Our approach is to present trust studies in three main steps. Introducing trust theories as first attempts to grasp notions linked to the concept of trust, fields of application, that explicit the uses that are traditionally associated to computational trust, and finally trust models, as an instantiation of a trust theory, w.r.t. some formal framework. Our survey ends with a set of issues that we deem important to deal with in priority in order to help the advancement of the field. The next two chapters present two models of trust. Our first model is an instantiation of Castelfranchi & Falcone's socio-cognitive trust theory. Our model is implemented using a Dynamic Epistemic Logic that we propose. The main originality of our solution is the fact that our trust definition extends the original model to complex action (programs, composed services, etc.) and the use of authored assignment as a special kind of atomic actions. The use of our model is then illustrated in a case study related to service-oriented architecture. Our second model extends our socio-cognitive definition to an abductive framework that allows us to associate trust to explanations. Our framework is an adaptation of Bochman's production relations to the epistemic case. Since Bochman approach was initially proposed to study causality, our definition of trust in this second model presents trust as a special case of causal reasoning, applied to a social context. We end our manuscript with a conclusion that presents how we would like to extend our work.

1 SUMMARY In general this thesis deals with the question whether or to what extent human thinking is rational in terms of the optimality of the way people achieve their goals and in terms of the consistency between people's beliefs and the structure of the world. This question is quite difficult to answer unequivocally because the answer will always depend on the nature of the particular task and the exact way in which we define rationality. Among other things, that's the reason why we can meet two contradictory schools of thought within the so-called Great Rationality Debate, one of which is convinced of the systematic irrationality of human thinking (in the sense of the systematic deviation of human thinking from normative predictions stemming from the principles of rational thinking as they are captured by the statistical theory of probability, formal logic or decision theory), while the other one considers human thought to be more or less rational, and finds the source of its (alleged) failure elsewhere. In the case of the latter, however, the question is how to explain the apparent existence of irrational behavior and interindividual differences in such behavior. One possible answer to this question is illustrated by Griffiths and Tenenbaum's Bayesian model of causal reasoning based on perceived...

No description available.

Les systèmes de vision artificielle sont aujourd'hui limités à l'extraction de données issues de ce que les caméras « voient ». Cependant, la compréhension de ce qu'elles voient peut être enrichie en associant la connaissance du contexte et la connaissance d'interprétation d'un humain.Dans ces travaux de thèse, nous proposons une approche associant des algorithmes de vision atificielle à une modélisation sémantique du contexte d'acquisition.Cette approche permet de réaliser un raisonnement sur la connaissance extraite des images par les caméras en temps réel. Ce raisonnement offre une réponse aux problèmes d'occlusion et d'erreurs de détections inhérents aux algorithmes de vision artificielle. Le système complet permet d'offrir un ensemble de services intelligents (guidage, comptage...) tout en respectant la vie privée des personnes observées. Ces travaux forment la première étape du développement d'un bâtiment intelligent qui peut automatiquement réagir et évoluer en observant l'activité de ces usagers, i.e., un bâtiment intelligent qui prend en compte les informations contextuelles.Le résultat, nommé WiseNET, est une intelligence artificielle en charge des décisions au niveau du bâtiment (qui pourrait être étendu à un groupe de bâtiments ou même a l'échelle d'un ville intelligente). Elle est aussi capable de dialoguer avec l'utilisateur ou l'administrateur humain de manière explicite. / To date, computer vision systems are limited to extract digital data of what the cameras "see". However, the meaning of what they observe could be greatly enhanced by environment and human-skills knowledge.In this work, we propose a new approach to cross-fertilize computer vision with contextual information, based on semantic modelization defined by an expert.This approach extracts the knowledge from images and uses it to perform real-time reasoning according to the contextual information, events of interest and logic rules. The reasoning with image knowledge allows to overcome some problems of computer vision such as occlusion and missed detections and to offer services such as people guidance and people counting. The proposed approach is the first step to develop an "all-seeing" smart building that can automatically react according to its evolving information, i.e., a context-aware smart building.The proposed framework, named WiseNET, is an artificial intelligence (AI) that is in charge of taking decisions in a smart building (which can be extended to a group of buildings or even a smart city). This AI enables the communication between the building itself and its users to be achieved by using a language understandable by humans.

[pt] Esta dissertação tem por objetivo apresentar os resultados da pesquisa de mestrado em que se procurou evidenciar algumas características da relação de David Hume com a indução. Segundo a interpretação corrente, Hume é o responsável por mostrar que nossa razão não é capaz de justificar qualquer um de nossos raciocínios indutivos. O problema de Hume também se caracteriza por ser um problema acerca da racionalidade da ciência, pois se seu método principal, a indução, não pode receber suporte racional, parece lícito afirmar que o resultado de uma inferência indutiva é irracional. A fim de delinear o campo exato em que se insere a crítica humeana, este texto irá mostrar como Hume apresenta suas teorias acerca do raciocínio causal em seu Tratado da natureza humana, traçar as características exatas do raciocínio causal de Hume e confrontá-las com as formas de interpretação presentes em alguns de seus principais comentadores. Procuramos tornar claras as falhas apresentadas nestas interpretações. Em seguida trataremos de discutir algumas das mais celebradas interpretações da filosofia de Hume, centrando nossa análise nos textos de Mackie, Beauchamp e Mappes. O capítulo final tem por objetivo mostrar as características racionais que podem ser atribuídas aos raciocínios causais humeanos, salientando o caráter particular de suas inferências. Finalizando, mostraremos como a origem do princípio da cópia pode ser um exemplo do uso de inferências indutivas por parte de Hume, o que nos leva a considerações heterodoxas a respeito de sua visão a respeito da racionalidade. / [en] The aim of this work is to present the results of my master´s degree research, which tried to show some of the characteristics of David Hume´s approach to induction. According to the standard interpretation, Hume is responsible for showing that our reason is not able to justify any of our inductive reasonings. Hume´s problem also characterizes itself by being a problem about the rationality of science, for, since his main method, induction, cannot receive a rational foundation, it seems licit to assert that the result of any inductive inference is irrational. In order to precisely describe the Humean criticism I am going to show how Hume presents his theories concerning causal reasoning in this A Treatise of Human Nature, define the exact characteristics of causal reasoning according to him, and compare this analysis to those by some of his main critics. We shall try to bring to light the proposed inadequacy of the latter. Next we will discuss some of the most celebrated interpretations of Hume´s philosophy, specially those by of Mackie, Beauchamp and Mappes. The final chapter aims at showing the rational characteristics that can be assigned to Humean causal reasoning emphasizing the particular character of his inferences. Finally, we show how the origin of the copy principle can be an instance of the use of inductive inferences by Hume, which allows us to risk some heterodox hypotheses concerning his view of rationality.

En aquesta dissertació es pretén valorar l'efectivitat de la WebQuest com a eina didàctica que afavoreix el desenvolupament dels processos d'ensenyament / aprenentatge de continguts curriculars propis de l'àrea de les Ciències Socials i la Història en l'etapa deEnsenyament Secundari Obligaotira. Per a això es realitza una àmplia revisió bibliogràfica sobre la recerca en WebQuests i es presenten 3 estudis quantitatius amb 210 alumnes en tres cursos de secundària, en què s'analitzen els sus textos (estudis 1 i 2) i el seu procés de resolució de l’activitat (estudi 3). Els resultats mostren diferències significatives per a la gran majoria de variables extretes de les anàlisis, a favor de l'experiència amb WebQuests. Es conclou que encara que les anàlisis estadístiques mostren un efecte favorable de l'experiència, convindrien estudis complementaris de tipus qualitatiu que expliquessin la naturalesa d'aquestes diferències d'una manera més detallada a fi de poder valorar amb més precisió l'efectivitat d'aquesta metodologia didàctica. / En esta disertación se pretende valorar la efectividad de la WebQuest como herramienta didáctica que favorecería del desarrollo de los procesos de enseñanza/aprendizaje de contenidos curriculares propios del área de las Ciencias Sociales y la Historia en la etapa de Enseñanza Secundaria Obligatoria. Para ello se realiza una amplia revisión bibliográfica sobre la investigación en WebQuests y se presentan 3 estudios cuantitativos con 210 alumnos en tres cursos de secundaria, en los que se analizan los textos que estos producen (estudios 1 y 2) y su proceso de resolución de la actividad (estudio 3). Los resultados muestran diferencias significativas para la gran mayoría de variables extraídas de los análisis, a favor de la experiencia con WebQuests. Se concluye que aunque los análisis estadísticos muestran un efecto favorable de la experiencia, convendrían estudios complementarios de tipo cualitativo que explicasen la naturaleza de estas diferencias de una manera más pormenorizada en orden a poder valorar con mayor precisión la efectividad de esta metodología didáctica. / This study assess the effectiveness of WebQuest as teaching and learning tools that enhance: 1) learning of contents specific to the area of Social Science and History, 2) causal reasoning and critical thinking development. To that end, a comprehensive literature review on WebQuests research is carried on, and 3 quantitative studies with 210 students in three secondary education levels is conducted by analyzing the texts students produce (Studies 1 and 2) and their performance level on the activity resolution process (Study 3). Results show significant differences for the vast majority of variables which means that the experience with WebQuests could enhance: 1) Social Science and History curricular contents learning, and 2) causal reasoning and critical thinking development. Despite statistical analysis showed a favorable effect of the experience, it is concluded that further qualitative additional studies would be helpful in order to explain the nature of these differences. These studies could spell out the way these differences are working, in order to better assess WebQuests effectiveness.