Avec l'apparition des paradigmes de l'intelligence ambiante et de la robotique ubiquitaire, on assiste à l'émergence de nouveaux systèmes intelligents ambiants visant à créer et gérer des environnements ou écosystèmes intelligents d'une façon intuitive et transparente. Ces environnements sont des espaces intelligents caractérisés notamment par l'ouverture, l'hétérogénéité, l'incertitude et la dynamicité des entités qui les constituent. Ces caractéristiques soulèvent ainsi des défis scientifiques considérables pour la conception et la mise en œuvre d'un système intelligent adéquat. Ces défis sont principalement au nombre de cinq : l'abstraction de la représentation des entités hétérogènes, la gestion des incertitudes, la réactivité aux événements, la sensibilité au contexte et l'auto-adaptation face aux changements imprévisibles qui se produisent dans l'environnement ambiant. L'approche par composition dynamique de services constitue l'une des réponses prometteuses à ces défis. Dans cette thèse, nous proposons un système intelligent capable d'effectuer une composition dynamique de services en tenant compte, d'une part, du contexte d'utilisation et des diverses fonctionnalités offertes par les services disponibles dans un environnement ambiant et d'autre part, des besoins variables exprimés par les utilisateurs. Ce système est construit suivant un modèle multicouche, adaptatif et réactif aux événements. Il repose aussi sur l'emploi d'un modèle de connaissances expressif permettant une ouverture plus large vers les différentes entités de l'environnement ambiant notamment : les dispositifs, les services, les événements, le contexte et les utilisateurs. Ce système intègre également un modèle de découverte et de classification de services afin de localiser et de préparer sémantiquement les services nécessaires à la composition de services. Cette composition est réalisée d'une façon automatique et dynamique en deux phases principales: la phase offline et la phase online. Dans la phase offline, un graphe global reliant tous les services abstraits disponibles est généré automatiquement en se basant sur des règles de décision sur les entrées et les sorties des services. Dans la phase online, des sous-graphes sont extraits automatiquement à partir du graphe global selon les tâches à réaliser qui sont déclenchées par des événements qui surviennent dans l'environnement ambiant. Les sous-graphes ainsi obtenus sont exécutés suivant un modèle de sélection et de monitoring de services pour tenir compte du contexte d'utilisation et garantir une meilleure qualité de service. Les différents modèles proposés ont été mis en œuvre et validés sur la plateforme ubiquitaire d'expérimentation du laboratoire LISSI à partir de plusieurs scénarii d'assistance et de maintien de personnes à domicile / With the appearance of the paradigms of the ambient intelligence and ubiquitaire robotics, we attend the emergence of new ambient intelligent systems to create and manage environments or intelligent ecosystems in a intuitive and transparent way. These environments are intelligent spaces characterized in particular by the opening, the heterogeneousness, the uncertainty and the dynamicité of the entities which establish(constitute) them. These characteristics so lift(raise) considerable scientific challenges for the conception(design) and the implementation of an adequate intelligent system. These challenges are mainly among five: the abstraction of the representation of the heterogeneous entities, the management of the uncertainties, the reactivity in the events, the sensibility in the context and the auto-adaptation

Im Rahmen der Dissertation wurde ein Middlewaredienst entwickelt und realisiert. Es gestattet die dynamische Verwaltung heterogener Kontextquellen. Das zugrunde liegende Komponentenmodell selbstbeschreibender Context Provieder ermöglicht die lose Kopplung von Kontextquellen und -senken. Es wird durch Filter- und Konverterkomponenten zur generischen Providersselektion anhand domänenspezifischer Merkmale ergänzt. Die Kopplung der verteilten Dienstinstanzen erfolgt durch ein hybrides Peer-to-Peer-System. Dies trägt der Heterogenität der Endgeräte Rechnung, und erlaubt die skalierbare , verteilte Verwaltung von Kontextquellen in globalen Szenarien.

In recent years pervasive computing has enjoyed an amazing growth in both research and commercial fields. Not only have the number of available techniques and tools expanded, but the number of actual deployments has been underwhelming. With this growth however, we are also experiencing a divergence of software interfaces, languages, and techniques. This leads to an understandably confusing landscape which needlessly burdens the development of applications. It is our sincere hope that through the use of specialized interfaces, languages, and tools, we can make pervasive computing environments more approachable and efficient to software developers and thereby increase the utility and value of pervasive computing applications. In this dissertation, we present a new method for creating and managing the long-term conversations between peers in pervasive computing environments. The Application Sessions Model formally describes these conversations and specifies techniques for managing them over their lifetimes. In addition to these descriptions, this dissertation presents a prototype implementation of the model and results from its use for realistic scenarios. To address the Application Sessions Model's unique needs for resource discovery in pervasive computing environments, we also present the Evolving Tuples Model. This model is also formally defined in this dissertation and practical examples are used to clarify its features. A prototype for both sensor hardware and software simulation of this model is described along with results characterizing the behavior of the model. The models, prototypes, and evaluations of both models presented here form the basis of a new and interesting line of research into support structures for pervasive computing application development. / text

Los Sistemas Perceptivos del Contexto son capaces de reconocer y anticipar proactivamente a necesidades próximas del usuario teniendo en cuenta intereses, ubicación y proximidad, incluso sin necesidad de comunicárselo de manera explícita. Esto gracias al aprovechamiento de las redes sociales como la herramienta de “Inteligencia de Negocios” que permitirá no sólo conectar a los consumidores con las marcas sino también para entender a los clientes finales y con ello desarrollar productos y servicios especializados. Context Awareness se apoya de los Sistemas de Posicionamiento Global (GSP) quienes facilitan servicios basados en la ubicación del usuario para la creación de soluciones contextuales. Diversas arquitecturas son propuestas para el desarrollo de soluciones contextuales sin embargo carecen de ubicuidad, lo que limita la aplicabilidad de las mismas. Este estudio propone una arquitectura lógica para un Sistema Perceptivo del Contexto que permita agilizar la fase de desarrollo reflejando los niveles de dependencia entre componentes hardware y software. A modo validación se proponen escenarios aplicativos que consumen los servicios y dominios propuestos en la arquitectura con ayuda del procesamiento de los sensores y dispositivos inteligentes del mercado móvil. De esta manera, el estudio mejora el desarrollo de los Sistemas Inteligentes. / Los Sistemas Perceptivos del Contexto son capaces de reconocer y anticipar proactivamente a necesidades próximas del usuario teniendo en cuenta intereses, ubicación y proximidad, incluso sin necesidad de comunicárselo de manera explícita. Esto gracias al aprovechamiento de las redes sociales como la herramienta de “Inteligencia de Negocios” que permitirá no sólo conectar a los consumidores con las marcas sino también para entender a los clientes finales y con ello desarrollar productos y servicios especializados. Context Awareness se apoya de los Sistemas de Posicionamiento Global (GSP) quienes facilitan servicios basados en la ubicación del usuario para la creación de soluciones contextuales. Diversas arquitecturas son propuestas para el desarrollo de soluciones contextuales sin embargo carecen de ubicuidad, lo que limita la aplicabilidad de las mismas. Este estudio propone una arquitectura lógica para un Sistema Perceptivo del Contexto que permita agilizar la fase de desarrollo reflejando los niveles de dependencia entre componentes hardware y software. A modo validación se proponen escenarios aplicativos que consumen los servicios y dominios propuestos en la arquitectura con ayuda del procesamiento de los sensores y dispositivos inteligentes del mercado móvil. De esta manera, el estudio mejora el desarrollo de los Sistemas Inteligentes.

Lors du développement des applications ubiquitaires, il est essentiel de définir des politiques de sécurité et de développerdes mécanismes de sécurité pour assurer la confidentialité et l’intégrité des applications. De nombreux langages de spécification de politiques se concentrent uniquement sur leur puissance d’expression. Les défis émergents dans les systèmes ubiquitaires ne peuvent pas être résolus par ces approches. Par exemple, la sensibilité au contexte est un élément central des systèmes ubiquitaires. Les approches existantes tiennent rarement compte des informations contextuelles dans leurs langages. Cette thèse propose une approche générative pour spécifier et implanter les politiques de sécurité dans les applications ubiquitaires. Pour définir une politique de sécurité, nous proposons un langage de spécification qui tient compte des informations contextuelles. Il permet aux développeurs de spécifier les règles de la politique et les entités requises (e.g. la description spatiale, les rôles, le contexte). Les politiques sont implémentés par des systèmes de réécriture, ce qui offre une grande puissance de vérification. Pour appliquer une politique, nous proposons une architecture qui intègre les concepts importants des politiques de sécurité (sujet, contexte, objet) dans des applications ubiquitaires. Pour mettre en oeuvre notre approche, nous avons enrichi une approche existante pour le développement des applications ubiquitaires. La spécification de la politique de sécurité et la description de l’application ubiquitaire enrichie sont utilisées pour générer un canevas de programmation qui facilite l’implémentation des mécanismes de sécurité, tout en séparant les aspects sécurités de la logique applicative. / When developing pervasive computing applications, it is critical to specify security policies and develop security mechanisms to ensure the confidentiality and integrity of the applications. Numerous policy specification languages only focus on their expressive power. The emerging challenges in pervasive computing systems can not be fulfilled by these approaches. For instance, context awareness is a central aspect of pervasive computing systems. Existing approaches rarely consider context information in their language.This thesis proposes a generative approach dedicated to specifying and enforcing security policies in pervasive computingapplications. To specify a policy, we propose a context-aware policy specification language which helps developers to specify policy rules and required entities (e.g. spatial description, roles, context information). Policies are implemented by term rewriting systems which offers great verification power. To enforce a policy, we propose an architecture that embeds important concepts of security policies (subject, object, security related context) into pervasive computing applications. To apply our approach, we enriched an existing approach which is dedicated to develop pervasive computing applications. Based on the policy specification and the enriched pervasive computing application descriptions, a dedicated programming framework is generated. This framework guides the implementation and raises the level of abstraction which can reduce the workloads of developers.

This study aimed to explore the underlying basis of the communication difficulties in children (between 5.0 and 7.11 years) with high functioning pervasive developmental disorder (HFPDD) (n=26), compared to children with specific language impairment (SLI) (n=26), and children with no history of developmental difficulty (NDD) (n=26). The study looked at: whether different profiles could be obtained for the groups on comprehensive batteries of communication, cognitive processing and theory of mind; which areas measured were best correlated; and which measures best differentiated the groups. Comprehensive communication and theory of mind batteries were devised and conducted. Cognitive processing was measured using the Cognitive Assessment System (CAS) (Naglieri and Das, 1997). Data was analysed using descriptive statistics, analysis of variance, the Kruskal-Wallis test, Bonferroni t tests, frequency distributions, Pearson correlation coefficients and discriminant function analyses. Kappa coefficients and analysis of variance measures were carried out on 23% of the rated data in order to establish inter-rater agreement and acceptable levels of agreement were reached. On the communication assessment, the HFPDD group experienced the most difficulty on the measures of understanding abstract vocabulary, understanding conversation, pronoun alternation, higher level semantics, narrative ability and pragmatic ability. In contrast, the SLI group experienced the most difficulty on the measures of expressive grammar and narrative clarity. On the cognitive processing assessment, the HFPDD group experienced marked difficulty in the areas of planning and attention, while the SLI group experienced significant difficulty in the areas of successive processing and less marked but still significant difficulty in the area of planning. Within the HFPDD group, a group with simultaneous processing markedly stronger than successive processing, a group with successive processing markedly stronger than simultaneous processing, and a group with simultaneous and successive processing occurring at a similar level, were identified. The HFPDD group experienced significant difficulty on all the measures of theory of mind, although a limited number of HFPDD subjects did not experience difficulty. The SLI group experienced significant difficulty on the two theory of mind measures that were more verbally loaded. Strong correlations were found between receptive language, expressive semantics, narrative ability, pragmatic ability, planning, attention and theory of mind; and between expressive grammar and successive processing. Pragmatic ability, narrative ability, planning, and certain of the theory of mind measures best appeared to discriminate the groups. A combined model of language, cognitive and theory of mind processing is proposed to explain the differences between the HFPDD and SLI groups.

Nos últimos anos, o volume de dados e de informações disponíveis pelas RSSF cresceu consideravelmente por ser uma tecnologia que permite o desenvolvimento de aplicações em diversos tipos de ambientes (pervasivos, ubíquos, ambientes controlados e de difícil acesso), fazendo com que a preocupação em gerenciar esse conteúdo se tornasse maior. Dentre as propostas apresentadas, um grande número utiliza ontologias como forma de organizar e categorizar dados; existem duas questões chaves na gerência de dados nas redes de sensores sem fio: o armazenamento de dados e o processamento da consulta. O presente trabalho propõe uma abordagem semântica para o gerenciamento dos dados dinâmicos e estáticos da rede, oferecendo um serviço de metadados sobre as RSSF e suas aplicações; esta técnica baseia-se na construção de ontologias que permitem escalabilidade, incremento de eficiência das atividades do ambiente e reusabilidade. Verificou-se mediante testes realizados baseados em consultas sobre a ontologia que o modelo proposto é apropriado, e que a busca baseada em semântica resulta na obtenção de dados das RSSF de forma mais adequada, quanto comparada com as abordagens tradicionais. / In the recent years, the volume of data available from wireless sensor networks (WSN) has increased considerably. WSN is a technology that allows the development of applications in various types of environments (pervasives, Ubiquitous, controlled and with difficult access environment), doing that concern for content managing is greater. Among the proposals, many use ontologies as a way to organize and categorize data; there are two key issues in data management in WSN: data storage and query processing. The present work proposes a semantic approach to the management of dynamic and static data from the network, providing a metadata service on WSN and its applications; this technique is based on the construction of ontologies that allows scalability, enhancement of efficiency in the activities of the environment and reusability. It was verified by means of tests based on querys on the ontology that the proposed model is appropriate, and that the search based on semantics provided retrieval of data from WSN more adequately, as compared with traditional approaches.

Introdução: A prevalência dos Transtornos Globais do Desenvolvimento (TGD) ou Transtornos Invasivos do Desenvolvimento (TID) tem sido descrita em 1% da população. A maioria dos estudos não sugere diferenças epidemiológicas entre as diversas culturas, entretanto, esses estudos têm sido realizados apenas na América do Norte e Europa. Para a realização de estimativas epidemiológicas sobre os TGD/TID, é importante que existam instrumentos devidamente validados em outras culturas. Objetivo: Tradução, retro-versão, adaptação cultural e validação da Autism Screening Questionnaire ou Social Communication Questionnaire, versão Lifetime, para a língua portuguesa e uso no Brasil. Método: Foi selecionada uma amostra inicial de 120 indivíduos, dividida em 3 grupos de 40 pacientes cada: pacientes com diagnóstico clínico de TGD/TID, com diagnóstico clínico de Síndrome de Down e com outros transtornos psiquiátricos. O questionário foi aplicado aos responsáveis legais dos indivíduos. Padrões de um questionário auto-aplicável foram seguidos. As medidas psicométricas do questionário, na sua versão final, foram testadas. Resultados: Valores de sensibilidade de 92,5% e especificidade de 95,5% foram encontrados para uma pontuação de 15. Esse escore teve valor discriminativo para os sujeitos com características de TGD/TID. A validade interna para o total de questões foi de 0,895 para o Alpha de Cronbach, e de 0,896 para o KR-20. Houve uma variação de 0,6 a 0,8 para ambos os coeficientes. Os valores de confiabilidade (Kappa) obtidos pelo teste e re-teste demonstraram que a maioria das questões obteve alta concordância. A análise fatorial com 3 fatores explicou 33,801% do total de variância. Conclusões: A versão final do instrumento de pesquisa traduzido e adaptado à cultura brasileira apresentou propriedades de medida satisfatórias, demonstrando ser de fácil aplicação, além de ser uma ferramenta útil para a realização de rastreamento diagnóstico em indivíduos com TGD/TID. / Introduction: The prevalence of pervasive developmental disorders (PDD) in world population is around 1%. Although studies do not suggest any modification of this prevalence in relation to ethnical differences, there are no studies describing the prevalence of PDD in other cultures but the northern hemisphere. It might be important to get efficient and validated diagnostic instruments in order to have that yet unknown prevalence. Objective: To translate into Portuguese, back-translate, culturally adapt and validate the Social Communication Questionnaire (Lifetime version), which is a screening instrument for pervasive developmental disorder (PDD), for use in Brazil. Method: A sample of 120 patients was selected and then divided in three different groups of 40. These groups were: patients with a clinical diagnosis of PDD, with Down Syndrome and with other psychiatric disorders. The selfreport questionnaire was applied to the patients legal guardians. Psychometric measures of the translated questionnaire final version were tested. Results: The score of 15 had a sensitivity of 92.5% and specificity of 95.0%. This same score was a cut-off point for the diagnosis of PDD. Internal validity for all 40 questions was 0.895 for Alpha, and 0.896 for KR-20, both coefficients ranging from 0.6 to 0.8. Test and retest reliability values (Kappa) showed that strong agreement in most of the questions. Conclusions: The final version of this instrument, which was translated into Portuguese and also adapted to Brazilian culture, had satisfactory measurement properties. It was an easy-to-apply and useful tool for the diagnostic screening of individuals with PDD

"Our modern society is increasingly relying on the collection, processing, and sharing of digital information. There are two fundamental trends: (1) Enabled by the rapid developments in sensor, wireless, and networking technologies, communication and networking are becoming more and more pervasive and ad hoc. (2) Driven by the explosive growth of hardware and software capabilities, computation power is becoming a public utility and information is often stored in centralized servers which facilitate ubiquitous access and sharing. Many emerging platforms and systems hinge on both dimensions, such as E-healthcare and Smart Grid. However, the majority information handled by these critical systems is usually sensitive and of high value, while various security breaches could compromise the social welfare of these systems. Thus there is an urgent need to develop security and privacy mechanisms to protect the authenticity, integrity and confidentiality of the collected data, and to control the disclosure of private information. In achieving that, two unique challenges arise: (1) There lacks centralized trusted parties in pervasive networking; (2) The remote data servers tend not to be trusted by system users in handling their data. They make existing security solutions developed for traditional networked information systems unsuitable. To this end, in this dissertation we propose a series of user-centric security and privacy mechanisms that resolve these challenging issues in untrusted network and computing environments, spanning wireless body area networks (WBAN), mobile social networks (MSN), and cloud computing. The main contributions of this dissertation are fourfold. First, we propose a secure ad hoc trust initialization protocol for WBAN, without relying on any pre-established security context among nodes, while defending against a powerful wireless attacker that may or may not compromise sensor nodes. The protocol is highly usable for a human user. Second, we present novel schemes for sharing sensitive information among distributed mobile hosts in MSN which preserves user privacy, where the users neither need to fully trust each other nor rely on any central trusted party. Third, to realize owner-controlled sharing of sensitive data stored on untrusted servers, we put forward a data access control framework using Multi-Authority Attribute-Based Encryption (ABE), that supports scalable fine-grained access and on-demand user revocation, and is free of key-escrow. Finally, we propose mechanisms for authorized keyword search over encrypted data on untrusted servers, with efficient multi-dimensional range, subset and equality query capabilities, and with enhanced search privacy. The common characteristic of our contributions is they minimize the extent of trust that users must place in the corresponding network or computing environments, in a way that is user-centric, i.e., favoring individual owners/users."

Submitted by Nara Lays Domingues Viana Oliveira (naradv) on 2015-07-10T17:40:18Z No. of bitstreams: 1 girola.pdf: 10642465 bytes, checksum: 2b0d8d06a973f586b2d20d3a4aa1e9ef (MD5) / Made available in DSpace on 2015-07-10T17:40:18Z (GMT). No. of bitstreams: 1 girola.pdf: 10642465 bytes, checksum: 2b0d8d06a973f586b2d20d3a4aa1e9ef (MD5) Previous issue date: 2014 / Nenhuma / Em um cenário de popularização de dispositivos móveis com alto poder de processamento e conectividade e de aumento na disponibilidade de redes sem fio e serviços em nuvem, é cada vez maior o número de aplicações distribuídas, nas mais diversas áreas. Esses sistemas precisam lidar com novos tipos de problemas advindos da natureza móvel e/ou ubíqua de seus nodos computacionais. Cada vez mais, os aplicativos em dispositivos móveis precisam examinar as informações de seu contexto e histórico de ações, a fim de permitir uma rápida adaptação, exigindo mínima distração do usuário e grande usabilidade. Além disso, o modelo cliente-servidor demonstra limitações para atender os requerimentos dos novos sistemas. Este trabalho apresenta o U-Deal, um modelo multiagente descentralizado para comércio ubíquo. O modelo descreve um sistema Peer-to-peer sensível ao contexto, em que se destaca um engine para análise de trilhas e um gerador de perfil dinâmico. O objetivo do modelo é suportar a recomendação automática de oportunidades de negócio, com base no perfil, preferências e histórico de atividades dos usuários. Neste trabalho, são apresentados uma especificação do sistema, utilizando a metodologia Prometheus, e um protótipo, incluindo uma avaliação através da execução de cenários simulados. / In a scenario of massive adoption of powerfull and highly connected mobile devices, cloud computing and improved wi-fi networks, is rapidly increasing the number of distributed application, in several areas. These systems need to handle new kind of problems, deriving from the mobile and/or ubiquitous nature of their computational nodes. Also, more and more mobile devices applications need to check context and historical information to adapt better to situations and provide minimum distraction and great usability to end user. Besides, the clientserver model shows limitations to handle the new systems requirements. This work presents the U-Deal, a decentralized and multi-agent ubiquitous commerce model, that describes a Peerto-peer context aware system, with a distributed historical context (trail) engine and a dynamic profile generator. The aim of the model is to provide an automatic deal opportunity recommendation based on user profile, preferences and activitiy history in a peer-to-peer environment. In this text, we present a system specification and a prototype, including simulated u-commerce scenarios.