Abstract Ubiquitous computing transforms physical environments into smart spaces, supporting users in an unobtrusive fashion. Such support requires sensing and interpreting the situation of the user, and providing the required functionality utilizing resources available. In other words, context acquisition, context modelling, and context reasoning are required. This thesis explores rule-based context reasoning from three perspectives: to implement the functionality of ubiquitous applications, to support the creation of ubiquitous applications, and to achieve self-adaptation. First, implementing functionality with reasoning is studied by comparing an application equipped with rule-based reasoning with an application providing similar functionality with hard coded application logic. The scalability of rule-based reasoning is studied with a large-scale student assistant scenario. Reasoning with constrained resources is explored with an application that performs reasoning partially on mobile devices. Finally, distributing a reasoning component that supports smart space interaction is explored with centralized, hybrid, and distributed architectures. Second, the creation of applications with rule-based reasoning is explored. In the first study, rules support building applications from available services and resources based on the instructions that users give via physical user interfaces. The second study supports developers, by proposing middleware that dynamically selects services and data based on the rules written by application developers. Third, self-adaptation is explored with a conceptual framework that adds self-introspective monitoring and control to smart space applications. This framework is verified with simulation and theoretical studies, and an application that fuses diverse data to provide fuel-efficient driving recommendations and adapts decision-making based on the driver’s progress and feedback. The thesis’ contributions include demonstrative cases on using rule-based reasoning from different perspectives, different scales, and with different architectures. Frameworks, a middleware, simulations, and prototypes provide the concrete contribution of the thesis. Generally, the thesis contributes to understanding how rule-based reasoning can be used in ubiquitous computing. The results presented can be used as guidelines for developers of ubiquitous applications. / Tiivistelmä Jokapaikan tietotekniikka muokkaa fyysisen ympäristömme älykkääksi tilaksi, joka tukee käyttäjää häntä häiritsemättä. Tuki toteutetaan asentamalla ympäristöön käyttäjää ja ympäristöä havainnoivia laitteita, tulkitsemalla kerätyn tiedon perusteella käyttäjän tilanne ja tarjoamalla tilanteeseen sopiva toiminnallisuus käyttäen saatavilla olevia resursseja. Toisin sanoen, älykkään tilan on kyettävä tunnistamaan ja mallintamaan toimintatilanne sekä päättelemään toimintatilanteen perusteella. Tässä työssä tutkitaan sääntöpohjaista päättelyä toimintatilanteen perusteella sovellusten toiminnallisuuden toteutuksen, kehittämisen tuen sekä mukautuvuuden näkökulmista. Sovellusten toiminnallisuuden toteuttamista päättelemällä tutkitaan vertaamalla sääntöpohjaisen päättelyn avulla toteutettua toiminnallisuutta vastaavaan suoraan sovellukseen ohjelmoituun toiminnallisuuteen. Sääntöpohjaisen päättelyn skaalautuvuutta arvioidaan laajamittaisessa opiskelija-assistenttiskenaariossa. Niukkojen resurssien vaikutusta päättelyyn arvioidaan päättelemällä osittain mobiililaitteessa. Älykkään tilan vuorovaikutusta tukevan päättelykomponentin hajauttamista tutkitaan keskitetyn, hybridi- ja hajautetun arkkitehtuurin avulla. Sovelluskehityksen tukemiseksi päättelyn säännöt muodostetaan saatavilla olevista palveluista ja resursseista käyttäjän fyysisen käyttöliittymän välityksellä antamien ohjeiden mukaisesti. Toisessa tapauksessa sovelluskehitystä tuetaan väliohjelmistolla, joka valitsee palvelut ja datan dynaamisesti sovelluskehittäjien luomien sääntöjen perusteella. Mukautuvuutta tutkitaan tilan hallintaan ja itsehavainnointiin liittyvän toiminnallisuuden lisäämiseen pystyvän käsitteellisen kehyksen avulla. Kehyksen toiminta varmennetaan simulointien sekä teoreettisten tarkastelujen avulla. Toteutettu useita datalähteitä yhdistävä sovellus antaa ajoneuvon kuljettajalle polttoaineen kulutuksen vähentämiseen liittyviä suosituksia sekä mukautuu kuljettajan ajotavan kehityksen ja palautteen perusteella. Työssä on osoitettu sääntöpohjaisen päättelyn toimivuus eri näkökulmista, eri skaalautuvuuden asteilla sekä eri arkkitehtuureissa. Työn konkreettisia tuloksia ovat kehykset, väliohjelmistot, simuloinnit sekä prototyypit. Laajemmassa mittakaavassa työ edesauttaa ymmärtämään sääntöpohjaisen päättelyn soveltamista ja työn tuloksia voidaankin käyttää suosituksina sovelluskehittäjille.

In a modern world software systems are literally everywhere. These should cope with very complex scenarios including the ability of context-awareness and self-adaptability. The concept of roles provide the means to model such complex, context-dependent systems. In role-based systems, the relational and context-dependent properties of objects are transferred into the roles that the object plays in a certain context. However, even if the domain can be expressed in a well-structured and modular way, role-based models can still be hard to comprehend due to the sophisticated semantics of roles, contexts and different constraints. Hence, unintended implications or inconsistencies may be overlooked. A feasible logical formalism is required here. In this setting Description Logics (DLs) fit very well as a starting point for further considerations since as a decidable fragment of first-order logic they have both an underlying formal semantics and decidable reasoning problems. DLs are a well-understood family of knowledge representation formalisms which allow to represent application domains in a well-structured way by DL-concepts, i.e. unary predicates, and DL-roles, i.e. binary predicates. However, classical DLs lack expressive power to formalise contextual knowledge which is crucial for formalising role-based systems. We investigate a novel family of contextualised description logics that is capable of expressing contextual knowledge and preserves decidability even in the presence of rigid DL-roles, i.e. relational structures that are context-independent. For these contextualised description logics we thoroughly analyse the complexity of the consistency problem. Furthermore, we present a mapping algorithm that allows for an automated translation from a formal role-based model, namely a Compartment Role Object Model (CROM), into a contextualised DL ontology. We prove the semantical correctness and provide ideas how features extending CROM can be expressed in our contextualised DLs. As final step for a completely automated analysis of role-based models, we investigate a practical reasoning algorithm and implement the first reasoner that can process contextual ontologies.

Submitted by Kamila Costa (kamilavasconceloscosta@gmail.com) on 2015-06-18T19:47:28Z No. of bitstreams: 1 Dissertação-Leonardo Lira Guimarães.pdf: 5378144 bytes, checksum: 17c3316e73590efa02a215df4d7e88b6 (MD5) / Approved for entry into archive by Divisão de Documentação/BC Biblioteca Central (ddbc@ufam.edu.br) on 2015-07-06T15:14:50Z (GMT) No. of bitstreams: 1 Dissertação-Leonardo Lira Guimarães.pdf: 5378144 bytes, checksum: 17c3316e73590efa02a215df4d7e88b6 (MD5) / Approved for entry into archive by Divisão de Documentação/BC Biblioteca Central (ddbc@ufam.edu.br) on 2015-07-06T15:21:38Z (GMT) No. of bitstreams: 1 Dissertação-Leonardo Lira Guimarães.pdf: 5378144 bytes, checksum: 17c3316e73590efa02a215df4d7e88b6 (MD5) / Made available in DSpace on 2015-07-06T15:21:38Z (GMT). No. of bitstreams: 1 Dissertação-Leonardo Lira Guimarães.pdf: 5378144 bytes, checksum: 17c3316e73590efa02a215df4d7e88b6 (MD5) Previous issue date: 2015-05-25 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Advances in wireless communication and computer hardware technologies have boosted the popularity of mobile devices. Increasingly, these devices gain new features of hardware (i.e., sensors and other gadgets) and software (e.g., facial, voice and gestures recognition) so that the human-computer interaction can occur more naturally. These features allowed a greater awareness of the environment and the conditions under which the users are, enabling the development of applications ever more proactive and sensitive. A context aware system can modify its behavior according to the inferred context of the environment. However, erroneous interpretations of the collected data may induce inappropriate and unwanted actions in applications. Although there is variety of inference techniques in the literature (e.g., rules, ontologies, that uses supervised and unsupervised learning), generally, they do not consider whether the inferences were indeed suitable to the user contexts. Furthermore, most of these techniques uses static inference models (i.e., they are unable to adjust themselves to changes in the environment conditions), which represents a limitation of these techniques when applied to the field of mobile applications. This work proposes a new context reasoning technique for mobile applications – called CoRe-RL – which uses reinforcement learning in order that the produced inferences could be ever more suitable to the user’s contexts. In this technique, learning occurs in an incremental manner and as the user interacts with the system, allowing the inference to be adjusted by the rewards (positive reinforcements) and punishments (negative reinforcements) associated to the inferred contexts. As the contexts are continuously being learned, the proposed technique also allows a flexible context management to the applications, which enables new contexts (labels) to be registered and learned over time. The operation of the technique is divided into two stages – classification and adaptation. The CoRe-RL uses a modified version of the K nearest neighbors in the classification stage. The learning (adaptation) stage is based on examples, but also makes adjustments on the models (features ranking) which weigh the most relevant xv features of each context in the classification stage. In order to validate and evaluate the proposed technique, it was developed, as a case study of this work, an application that implements all of the functionality and capabilities of CoRe-RL. Through this application, practical experiments for evaluating the classification and adaptation were executed in two specific scenarios: there was a single context in the first scenario; and in the second, there were three. Through the practical experiments, it was observed that, in accordance to the cutting threshold used, it is possible to obtain good performances in the classification even with a small base and with a slightly adjusted ranking. Furthermore, it was demonstrated that the CoRe-RL improves its performance, converging to the optimal performance, in accordance to the occurrence of new interactions. / Os avanços das tecnologias de comunicação sem fio e de hardware impulsionaram a popularização do uso de dispositivos móveis. Cada vez mais, estes dispositivos ganham novos recursos de hardware (i.e., sensores e outros gadgets) e software (e.g., reconhecimento facial, de voz, gestos) a fim de que a interação humano-computador ocorra de forma mais natural. Esses recursos deram aos dispositivos móveis uma capacidade maior de percepção do ambiente e das condições nas quais os usuários se encontram, possibilitando o desenvolvimento de aplicações cada vez mais proativas e sensíveis ao contexto. Um sistema sensível ao contexto é capaz de modificar seu comportamento de acordo com os contextos inferidos do ambiente. Entretanto, interpretações errôneas dos dados coletados podem induzir ações inapropriadas e indesejadas nas aplicações. Embora exista uma variedade de técnicas de inferência na literatura (e.g., regras, ontologias, que utilizam aprendizagem supervisionada e não supervisionada), em geral, elas não consideram se as inferências foram de fato adequadas para os contextos do usuário. Além disso, a maioria dessas técnicas utiliza modelos estáticos de inferência (i.e., que não são capazes de se ajustar à mudanças nas condições do ambiente), o que representa uma limitação dessas técnicas quando aplicadas ao domínio das aplicações móveis. Neste trabalho, é proposta uma nova técnica de inferência de contexto para aplicações móveis – chamada de CoRe-RL – que utiliza aprendizagem por reforço a fim de que sejam produzidas inferências cada vez mais adequadas aos contextos do usuário. Nesta técnica, a aprendizagem ocorre de maneira incremental e conforme o usuário interage com o sistema, permitindo que a inferência seja ajustada por meio de recompensas (reforços positivos) e punições (reforços negativos) associadas aos contextos inferidos. Como os contextos estão continuamente sendo aprendidos, a técnica proposta também permite às aplicações um gerenciamento flexível de contextos, ou seja, é possível que novos contextos (rótulos) sejam cadastrados e aprendidos ao longo do tempo. O funcionamento da técnica é divido em duas etapas – classificação e adapxiii tação. O CoRe-RL utiliza o método dos K vizinhos mais próximos (modificado) na classificação. A aprendizagem (adaptação) é baseada em exemplos, mas também faz ajustes sobre os modelos (ranking de características) que ponderam as características mais relevantes de cada contexto, na etapa de classificação. Com o intuito de testar e avaliar o desempenho da técnica proposta, foi desenvolvido, como estudo de caso deste trabalho, um aplicativo que implementa todas as funcionalidades e recursos do CoRe-RL. Através deste aplicativo, foram realizados experimentos práticos de avaliação da classificação e adaptação, em dois cenários específicos: no primeiro cenário havia um único contexto; e no segundo haviam três. Por meio dos experimentos práticos, observou-se que, de acordo com o limiar de corte usado, é possível obter bons desempenhos na classificação mesmo com uma base pequena e com um ranking pouco ajustado. Além disso, demonstrou-se que o CoRe-RL melhora seu desempenho, convergindo para o desempenho ótimo, de acordo com a ocorrência das interações.

Recent advances in technology are changing the way how everyday activities are performed. Technologies in the traffic domain provide diverse instruments of gathering and analysing data for more fuel-efficient, safe, and convenient travelling for both drivers and passengers. In this article, we propose a reference architecture for a context-aware driving assistant system. Moreover, we exemplify this architecture with a real prototype of a driving assistance system called Driving coach. This prototype collects, fuses and analyses diverse information, like digital map, weather, traffic situation, as well as vehicle information to provide drivers in-depth information regarding their previous trip along with personalised hints to improve their fuel-efficient driving in the future. The Driving coach system monitors its own performance, as well as driver feedback to correct itself to serve the driver more appropriately.

Context-aware computing aims at making mobile devices sensitive to the social and physical settings in which they are used. A necessary requirement to achieve this goal is to enable those devices to establish a shared understanding of the desired settings. Establishing a shared understanding entails the need to manipulate sensed data in order to capture a real world situation wholly, conceptually, and meaningfully. Quite often, however, the data acquired from sensors can be inexact, incomplete, and/or uncertain. Inexact sensing arises mostly due to the inherent limitation of sensors to capture a real world phenomenon precisely. Incompleteness is caused by the absence of a mechanism to capture certain real-world aspects; and uncertainty stems from the lack of knowledge about the reliability of the sensing sources, such as their sensing range, accuracy, and resolution. The thesis identifies a set of criteria for a context-aware system to capture dynamic real-world situations. On the basis of these criteria, a distributed architecture is designed, implemented and tested. The architecture consists of Primitive Context Servers, which abstract the acquisition of primitive contexts from physical sensors; Aggregators, to minimise error caused by inconsistent sensing, and to gather correlated primitive contexts pertaining to a particular entity or situation; a Knowledge Base and an Empirical Ambient Knowledge Component, to model dynamic properties of entities with facts and beliefs; and a Composer, to reason about dynamic real-world situations on the basis of sensed data. Two additional components, namely, the Event Handler and the Rule Organiser, are responsible for dynamically generating context rules by associating decision events ? signifying a user?s activity ? with the context in which those decision events are produced. Context-rules are essential elements with which the behaviour of mobile devices can be controlled and useful services can be provided. Four estimation and recognition schemes, namely, Fuzzy Logic, Hidden Markov Models, Dempster-Schafer Theory of Evidence, and Bayesian Networks, are investigated, and their suitability for the implementation of the components of the architecture of the thesis is studied. Subsequently, fuzzy sets are chosen to model dynamic properties of entities. Dempster-Schafer?s combination theory is chosen for aggregating primitive contexts; and Bayesian Networks are chosen to reason about a higher-level context, which is an abstraction of a real-world situation. A Bayesian Composer is implemented to demonstrate the capability of the architecture in dealing with uncertainty, in revising the belief of the Empirical Ambient Knowledge Component, in dealing with the dynamics of primitive contexts and in dynamically defining contextual states. The Composer could be able to reason about the whereabouts of a person in the absence of any localisation sensor. Thermal, relative humidity, light intensity properties of a place as well as time information were employed to model and reason about a place. Consequently, depending on the variety and reliability of the sensors employed, the Composer could be able to discriminate between rooms, corridors, a building, or an outdoor place with different degrees of uncertainty. The Context-Aware E-Pad (CAEP) application is designed and implemented to demonstrate how applications can employ a higher-level context without the need to directly deal with its composition, and how a context rule can be generated by associating the activities (decision events) of a mobile user with the context in which the decision events are produced.

Submitted by Cássia Santos (cassia.bcufg@gmail.com) on 2015-10-23T12:11:54Z No. of bitstreams: 2 Dissertação - Guilherme Melo e Maranhão - 2015.pdf: 2879073 bytes, checksum: 25e096adcb3251cd2153a4b4c5e41a3a (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Approved for entry into archive by Luciana Ferreira (lucgeral@gmail.com) on 2015-10-23T13:29:49Z (GMT) No. of bitstreams: 2 Dissertação - Guilherme Melo e Maranhão - 2015.pdf: 2879073 bytes, checksum: 25e096adcb3251cd2153a4b4c5e41a3a (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) / Made available in DSpace on 2015-10-23T13:29:49Z (GMT). No. of bitstreams: 2 Dissertação - Guilherme Melo e Maranhão - 2015.pdf: 2879073 bytes, checksum: 25e096adcb3251cd2153a4b4c5e41a3a (MD5) license_rdf: 23148 bytes, checksum: 9da0b6dfac957114c6a7714714b86306 (MD5) Previous issue date: 2015-05-06 / Context-aware computing investigates how to seamlessly enable the interaction among computing devices and the environment, or context, in which they are located in. The literature has defined the life cycle of context data as composed of four phases: acquisition, modelling, reasoning and dissemination. These latter two phases have been strongly influenced by the heterogeneity of published data, as a consequence of the increasing deployment of sensors. A great challenge reported in the literature has been the development of mechanisms for handling and reasoning about sensor data, and also for notifying contextaware applications in an efficient and relevant manner. Aiming to solve this problem, this research focuses on a new approach for semantic context reasoning and filtering in accordance with design principles recommended by the literature. As a result of this work, the main contributions include an extensible and maintainable mechanism for semantic context filtering; the architectural model of a context reasoning component, called Hermes Interpreter (HI); the Hermes Base component, which exposes an API for accessing a publish-subscribe-based communication middleware; and HI’s functional validation and performance evaluation in a simulated scenario of vital signs monitoring in Intensive Care Units and wards. This research demonstrated the efficiency of the behaviour of the semantic filtering mechanism for end users applications. Besides, by increasing the number of subscribers, it was observed that the response time was acceptable in almost all experiments. Despite this, it was also verified the high cost of the semantic filtering processing in comparison with pure context reasoning activities. Regarding the HI component’s architectural design, this work recommends it as a reusable artifact for researches on the subject of context reasoning. / A computação sensível a contexto é a área da Ciência da Computação que estuda os meios para possibilitar a interação entre dispositivos computacionais e o ambiente, ou contexto, em que estão inseridos. Para promover essa interação, a literatura define quatro etapas para o ciclo de vida da informação de contexto: aquisição, modelagem, interpretação e disseminação. Dentre essas etapas, as de interpretação e disseminação têm sido fortemente impactadas pela heterogeneidade dos dados publicados, resultante da crescente implantação de sensores. Criar mecanismos para manipular e interpretar essas informações e depois notificá-las de maneira eficiente e com relevância para as aplicações sensíveis a contexto tem sido um desafio reportado pela ciência. Diante do problema apresentado, esse trabalho objetiva alinhar os princípios de projeto recomendados pela literatura com uma proposta de abordagem para interpretação e filtragem de contexto fundamentados em modelagem contextual semântica. Portanto, as principais contribuições deste trabalho incluem um mecanismo extensível e manutenível de filtragem semântica de contexto; o modelo arquitetural de um componente interpretador de contexto, denominado Hermes Interpreter, desenvolvido a partir de princípios de projeto defendidos da literatura; o componente Hermes Base, que expõe API de acesso para um middleware de comunicação, que opera sob o paradigma publish/subscribe; e por último, validações funcionais e de desempenho do Hermes Interpreter em cenário de simulação de monitoramento de sinais vitais de pacientes em UTIs e enfermarias. Por meio delas, comprovou-se a eficiência do mecanismo de filtragem para notificar de maneira precisa e eficiente os dados relevantes para os usuários. Além disso, observou-se que, devido a características de implementação, por maior que seja a quantidade de assinantes, o tempo de resposta foi aceitável em praticamente todos os experimentos. Apesar disso, constatou-se o alto custo do processamento de filtros em comparação às atividades puramente de interpretação de contexto. Por último, o projeto arquitetural do componente constitui um artefato reutilizável para pesquisas na área de intepretação de contexto.

Context-aware computing aims at making mobile devices sensitive to the social and physical settings in which they are used. A necessary requirement to achieve this goal is to enable those devices to establish a shared understanding of the desired settings. Establishing a shared understanding entails the need to manipulate sensed data in order to capture a real world situation wholly, conceptually, and meaningfully. Quite often, however, the data acquired from sensors can be inexact, incomplete, and/or uncertain. Inexact sensing arises mostly due to the inherent limitation of sensors to capture a real world phenomenon precisely. Incompleteness is caused by the absence of a mechanism to capture certain real-world aspects; and uncertainty stems from the lack of knowledge about the reliability of the sensing sources, such as their sensing range, accuracy, and resolution. The thesis identifies a set of criteria for a context-aware system to capture dynamic real-world situations. On the basis of these criteria, a distributed architecture is designed, implemented and tested. The architecture consists of Primitive Context Servers, which abstract the acquisition of primitive contexts from physical sensors; Aggregators, to minimise error caused by inconsistent sensing, and to gather correlated primitive contexts pertaining to a particular entity or situation; a Knowledge Base and an Empirical Ambient Knowledge Component, to model dynamic properties of entities with facts and beliefs; and a Composer, to reason about dynamic real-world situations on the basis of sensed data. Two additional components, namely, the Event Handler and the Rule Organiser, are responsible for dynamically generating context rules by associating decision events ? signifying a user?s activity ? with the context in which those decision events are produced. Context-rules are essential elements with which the behaviour of mobile devices can be controlled and useful services can be provided. Four estimation and recognition schemes, namely, Fuzzy Logic, Hidden Markov Models, Dempster-Schafer Theory of Evidence, and Bayesian Networks, are investigated, and their suitability for the implementation of the components of the architecture of the thesis is studied. Subsequently, fuzzy sets are chosen to model dynamic properties of entities. Dempster-Schafer?s combination theory is chosen for aggregating primitive contexts; and Bayesian Networks are chosen to reason about a higher-level context, which is an abstraction of a real-world situation. A Bayesian Composer is implemented to demonstrate the capability of the architecture in dealing with uncertainty, in revising the belief of the Empirical Ambient Knowledge Component, in dealing with the dynamics of primitive contexts and in dynamically defining contextual states. The Composer could be able to reason about the whereabouts of a person in the absence of any localisation sensor. Thermal, relative humidity, light intensity properties of a place as well as time information were employed to model and reason about a place. Consequently, depending on the variety and reliability of the sensors employed, the Composer could be able to discriminate between rooms, corridors, a building, or an outdoor place with different degrees of uncertainty. The Context-Aware E-Pad (CAEP) application is designed and implemented to demonstrate how applications can employ a higher-level context without the need to directly deal with its composition, and how a context rule can be generated by associating the activities (decision events) of a mobile user with the context in which the decision events are produced.