Recent trends and developments in the fields of Big Data, Machine Learning, and Artificial Intelligence are completely transforming the way brands are engaging and communicating with their audience, allowing more personalized communications than ever. With the spread of social networking sites, such as Facebook, YouTube or Instagram a new opportunity arises for companies to connect to their consumers. However, since social media personalization implies the collection and analysis of highly personal data, consumers may develop negative reactions, attitudes or perceptions towards personalized advertising. Research covers extensively issues such as privacy concerns, invasiveness, forced exposure or irritation, which can lead to advertising avoidance. Though understanding the user perspective is crucial, the topic advertising avoidance in the context of personalization, especially in social media environments, hasn’t been discussed at great length. The purpose of this thesis is to understand how YouTube users experience personalized advertising while using the platform. The empirical findings of this thesis contribute to the ongoing research on personalized advertising within a social media setting. In addition, by understanding what can influence personalized ad avoidance on YouTube and describing how consumers express ad avoidance on the platform, this thesis aims to nurture a deeper understanding of the phenomenon. This study is based on an interpretative, abductive as well as a qualitative research design. It uses semi-structured interviews to explore the views, experiences, beliefs, and motivations of individual participants as well as focus groups that can leverage group dynamics to generate new qualitative data. The results of this thesis show that YouTube users experience ad avoidance in relationship with personalized ads for several reasons linked to prior negative attitudes, perceived goal impediment or ad irritation. The analysis of the findings revealed that users can experience cognitive, affective and behavioral ad avoidance as presented in the literature, but a theoretical model which can perfectly explain the phenomena of personalized ad avoidance on social media is currently not existent. While some antecedents claimed by the previously mentioned theoretical frameworks were also visible in the study, additional aspects may have an impact on how consumers experience personalized advertising avoidance in the social media environment, and more specifically on YouTube.

Media usage, as well as media advertising, is growing in tandem with each other. Media users will, regardless of which medium that is in use, encounter a considerable amount of advertising. As we are moving towards a more digitized world, and also a more digitized media sphere, a new communication landscape has been formed. The digitalization has provided new ways to communicate and also new ways to advertise. Now there is not only possible to advertise through traditional offline media channels, such as radio and television, but also possible to advertise through online media, such as Facebook and Instagram. Personalized advertising is an advertising format which has evolved with the growth of online advertising and the ability to collect media users’ data online. Personalized advertisements are formed to comply with the media users’ preferences and have been proven to be effective and in some cases appreciated. Though, this form of advertisements has brought up concerns regarding the media users’ integrity online and previous research does, therefore, claim that personalized advertising is a paradox. Furthermore, a question that can be asked regarding advertising and the new communication landscape that has emerged is: what can be said about media users’ attitudes towards advertising online compared to advertising offline? The purpose of this study is to examine media users’ attitudes towards advertising online, compared to attitudes towards advertising offline, and further investigate whether the mediums themselves seem to have an impact on the receiver’s attitudes towards advertising. The study’s three research questions compose the structure which is followed throughout the study. Furthermore, to answer the research questions, a survey has been implemented. The survey intended to investigate media users’ attitudes towards advertisements online and offline. The theories the study is based on are Marshall McLuhan’s theory The Medium is the Message, The advertising value model by Ducoffe and Personalization. Based on the study’s theory and theoretical framework, the result was analyzed and discussed. The result of the study shows that media users’ attitudes towards advertising are overall rather similar, regardless of age, gender and political position. Further, a pattern that shows connections between media users’ self-reported awareness, their consent of data collection and their appreciation of personalized advertising online was found. The study further examines differences between attitudes towards different mediums, however, we can conclude that the mediums themselves have been proven to not be determinant in media users’ attitudes towards advertising - the experiences of advertising are rather negative regardless of where the media user encounters them. This study contributes with information and insight regarding media users’ attitudes towards advertising in different mediums, further, how media users stand regarding online privacy and concession with data collection in relation to their attitudes towards personalized advertising.

La réalisation d'une mission robotique se décompose usuellement en trois étapes: la planification, i.e. le choix des taches à réaliser, le séquencement, i.e. la détermination du timing et de l'ordre de réalisation des tâches, et finalement l'exécution du plan de tâches. Pour les systèmes redondants tels que les robots humanoïdes, la tâche (dans le sens de fonction de tâche) détermine une commande sur une partie du robot, permettant ainsi la réalisation simultanée de plusieurs tâches à l'aide d'un formalisme de pile de tâches. Cependant, les mécanismes d'ordonnancement classiques ne gèrent pas les cas où le mouvement est déterminé par un ensemble de tâches hiérarchisé: pour ces robots, la phase d'ordonnancement est éludée et l'exécution se base directement sur la plan de tâches donné par le planificateur. Le but de cette thèse est de réintroduire cette phase d'ordonnancement, tout en maintenant le rôle central de la tâche. Dans un premier temps, la continuité de la commande fournie par la pile de tâches est étudiée. En particulier, nous mettons en évidence les discontinuités accompagnant la réalisation d'événements discrets (à savoir l'insertion, le retrait et l'échange de priorité de tâches), puis proposons et comparons plusieurs méthodes de lissage. Ensuite, nous présentons une méthode permettant d'optimiser une séquence de tâches donnée en modifiant le timing et la paramétrisation des tâches, tout en respectant les contraintes liées à l'environnement. Enfin, une nouvelle utilisation de la flexibilité de la fonction de tâche consistant à adapter une séquence de tâches aux préférences d'un utilisateur humain est illustrée. Ces résultats sont illustrés sur un robot humanoïde. / A general agreed approach on mission and motion planning consists in splitting it into three steps: decomposing the mission into a sequence of tasks (task planning), determining the order of realization and the timing of the tasks (task scheduling) and finally executing the task sequence. This approach maintains the task component in the entire reasoning, using it as a bridge between planning, scheduling and execution.In the sense of task function, a task defines a control law on part of the robot. Hence, for highly redundant systems such as humanoid robots, it is possible to realize several tasks simultaneously using a stack-of-tasks formalism. Though, classical schedulers do not handle the case where the motion is specified not by one, but by a combination of tasks organized into a hierarchy. As a result, the scheduling phase is usually skipped. This thesis aims at reintroducing the scheduling phase, while maintaining the central role of the task.First, the stack-of-tasks formalism is recalled and the continuity of the control law is studied. Particularly, we show that discreet operations (insertion, removal and swap of priority between tasks) create discontinuities in the control. We then present and discuss smoothing methods. Second, we present a task-overlapping based method to optimize not only the scheduling but also the behavior of the tasks of a given sequence, while accounting for the physical constraint of the execution. Finally, we introduce a new perspective in the usage of the task-function approach the task function approach to personalize a task sequence and take into account user preferences.These results are experimented on the humanoid robot platform HRP-2.

xiii, 80 p. : ill. (some col.) / The design of successful assistive technologies requires careful personalization for individual users, as well as rapid, low cost cycles for product development and testing. My research brings two modern software engineering models to meet these challenges: Personal and Contextual Requirements Engineering (PC-RE) and Agile Software Development. We adapt these models to the assistive mobile navigation domain for the blind. This dissertation demonstrates that a Virtual Environment testing can significantly reduce testing time, yield meaningful testing results by fully controlling environmental variables, alleviate logistical and safety problems, and serve as an ideal platform for deep personalization. We developed a narrative NAvigation Virtual Environment (NAVE) and compared blind subjects' performance and behavior in wayfinding tasks with tactile maps under field testing versus testing in NAVE. Our experiments showed positive results to support our hypothesis that virtual environments can be useful in replacing field testing for personalized assistive technologies in agile development. / Committee in charge: Stephen Fickas， Chairperson； Art Farley， Member； Michal Young， Member； Amy Lobben， Outside Member

O fenômeno global de envelhecimento populacional está sendo acompanhado pelo rápido crescimento do número de pessoas com algum tipo de deficiência mental ou física, assim como diversas doenças crônicas relacionadas à idade. Devido ao fato de que a maioria dos idosos prefere ficar no conforto de suas residências, é essencial o desenvolvimento de tecnologias que auxiliem os idosos a terem uma vida independente e confortável em seus lares. Em especial, Ambientes Assistidos buscam agir proativamente de forma a auxilar de maneira correta o usuário na execução das suas Atividades de Vida Diária. Devido à grande variabilidade nos níveis de deficiência dos usuários, é fundamental que o Ambiente Assistido seja capaz de fornecer a assistência levando em consideração diversos aspectos relacionados ao usuário, como suas dificuldades, limitações e atividade sendo realizada. Este trabalho propõe um sistema que utiliza uma abordagem baseada nas tecnologias da Web Semântica e no paradigma multiagentes com o objetivo de fornecer assistência personalizada a pessoas idosas e com diversidades funcionais. As características do usuário foram modeladas por uma ontologia OWL, enquanto o mecanismo de reasoning responsável por inferir as preferências relacionadas aos serviços do ambiente foi implementada utilizando a linguagem SWRL. Ainda, os componentes presentes na arquitetura proposta são implementados na forma de agentes com funções específicas. Entre as característica inovadoras da proposta estão a incorporação de um modelo padronizado e internacionalmente conhecido para representar as deficiências, dificuldades e limitações dos usuários e a utilização de uma arquitetura distribuída que possui as características inerentes ao paradigma multiagentes, tais como escalabilidade, flexibilidade e dinamismo. A validação do sistema foi realizada através de sua integração com uma Arquitetura Orientada a Serviços que utiliza uma solução comercial de automação residencial e um simulador de Ambientes Inteligentes, e da sua utilização em três casos de usos definidos pelo projeto AAL Joint Programme. / The global phenomenon of population aging has been followed by the fast growth in the number of people with some kind of mental or physical disability and chronic illnesses associated with old age. Given the fact that the majority of older adults prefer to stay in the comfort of their own homes, it is crucial to develop technologies that help the elderly to live independently and comfortably in their homes. Specifically, Ambient Assisted Living environments aims to act proactively in order to properly assist the elderly and people with functional diversity to perform their Activities of Daily Living. Since these users tend to be greatly variable in regards to functioning and disability levels, it is essential that the ambient provides personalized assistance based on the user’s unique requirements, limitations, preferences, and activity being performed. This work proposes a system that uses an approach based on the Web Semantic technologies and the multiagent paradigm with the goal of providing personalized assistance to the elderly and people with functional diversity. The user’s characteristics were modeled using an OWL ontology and a reasoning mechanism responsible to infer the preferences related to the ambient services was implemented using SWRL rules. Furthermore, the proposed architecture components were implemented as agents with specific roles. Among the proposed system innovative features are the incorporation of a standardized and international model used to represent the user’s functional and disability information, and the use of a distributed architecture having the multiagent paradigm features, such as scalability, flexibility, and dynamism. The system’s validation was accomplished by its integration with a Service Oriented Architecture that uses a commercial automation solution and a smart environment simulator, and by its use in three use cases developed by the AAL Joint Programme.

As melhorias alcançadas na Educação através do uso das tecnologias de informação e comunicação são inegáveis. Entretanto, os avanços em direção a um ensino personalizado e direcionado às preferências, ao nivelamento (saneamento das dificuldades e pré-requisitos de conteúdo) e aos interesses dos estudantes são pequenos. Desta forma, através da identificação dos conhecimentos prévios dos alunos, bem como dos conteúdos (conceituais) buscou-se aprimorar o processo de ensino para que cada aluno alcance, a seu modo (de forma personalizada) o aprendizado e possa avançar na disciplina/ano em questão. Como é humanamente impossível um professor conhecer as características e dificuldades de conteúdo de todos os seus alunos, é interessante que o ambiente virtual que apoia o professor possa auxiliar para um atendimento mais personalizado, permitindo que o aluno interaja no seu próprio ritmo de estudo, acesse os conteúdos necessários ou aprofunde-se em algum conteúdo que ele já domine. Esta tese propõe um método de uso das ferramentas disponíveis de ambientes virtuais, principalmente do MOODLE, para alcançar a personalização do ensino e conduzir o aluno em sua própria trajetória individual rumo ao aprendizado significativo, sanando dificuldades de conteúdo. Seguindo o objetivo de melhorar a qualidade do ensino sem necessidade de grade investimento de desenvolvimento de ambiente nem aporte financeiro. Para validar o método foi realizado um projeto piloto mais dois experimentos, todos no Colégio Militar de Porto Alegre que totalizaram a participação de 256 alunos. Os resultados demonstraram a viabilidade do método, medido através da melhora nas notas (análise quantitativa) e da análise qualitativa dos dados e dos questionários (respondidos pelos alunos e professores). Destaca-se como vantagem educacional o desenvolvimento do método que pode ser aplicado a qualquer disciplina ou modelo de ensino e como vantagem computacional o uso do MOODLE para personalizar o ensino. / The use of information and communication technologies made undeniable advances in education. However, advances towards personalized learning and directed to the preferences, the leveling (improvement of difficulties and prerequisites) and to the interests of students are small yet. Thus, it is intended, by identifying the students' prior knowledge and content (conceptual) to improve the process so that each student reach, at their mode (personalized) the learning and make progress in discipline/year in question. We know that is humanly impossible for a teacher to know the characteristics and difficulties of content of all his students. So, the virtual environment that supports the teacher should have the flexibility to customize interactions with each student, which allows the student to interact at his/her own study pace, access the necessary content or deepen in some content that he/she has mastered. This thesis proposes a method for using the available virtual environments tools, especially the MOODLE, to achieve the personalization of teaching and to lead the student in his own individual trajectory toward meaningful learning, healing content difficulties. Following the objective of improving the quality of education without the need for large investment in environmental development or financial contribution. To validate the method was carried out a pilot project plus two experiments, all occurred at Military College of Porto Alegre, with a total of 256 students. The results showed the viability of the method, measured by the improvement in grades (quantitative analysis) and the qualitative analysis of data and questionnaires (answered by students and teachers). It is highlighted as an educational advantage, the development of the method that can be applied to any discipline or teaching model and as computational advantage, the use of MOODLE to customize teaching.

Cette thèse s'intéresse à la personnalisation et à la personnification d'agents intelligents dans le cadre d'applications web. Les techniques de personnalisation et de personnification sont de plus en plusutilisées pour mieux répondre aux besoins des utilisateurs. La plupart de ces techniques reposent sur des outils de raisonnement issus de l'intelligence artificielle. Cependant, ces techniques sont habituellement utilisés de manière ad-hoc pour chaque application. L'approche de cette thèse est de considérer la personnalisation et la personnification comme deux instances d'altération de comportement, et à étudier l'altération du comportements d'agents rationnels. Les principales contributions sont WAIG, un formalisme basé sur le langage Golog adapté à l'expression d'applications web, et PAGE, un cadre formel pour la manipulation et l'altération de programmes agents Golog, permettant la transformation automatique d'agent selon un critère donné. Ces contributions sont illustrés par plusieurs scénarios concrets issus des domaines de la personnalisation et de la personnification. / This thesis deals with personalization and personification of rational agents within the framework of web applications. Personalization and personification techniques are more and more used to answer the needs of users. Most of those techniques are based on reasoning tools that come from the artificial inteligence field. However, those techniques are usually used in an ad-hoc way for each application. The approach of this thesis is to consider personaliaation and personification as two instances of alteration of behaviour, and to study the alteration of the behaviours of rational agents. The main contributions are WAIG, a formalism for the expression of web applications based on the agent programming language Golog, and PAGE, a formal framework for the manipulation and the alteration of Golog agent programs, which allow to transform an agent automatically following a given criterion. Those contributions are illustrated by concrete scenarios from the fields of personalization and personification.

This thesis investigated how clients and therapists discuss the means by which clients can work towards their therapeutic goals. Cooper et al. (2016) termed such discussions meta-therapeutic communication or meta-therapeutic dialogue and Cooper and McLeod (2011) recommend carrying them out since outcomes are robustly related to whether the client accepts the therapeutic strategy as appropriate for their needs (e.g. Horvath et al., 2011). This thesis undertook the first discovery-oriented, Conversation Analysis (CA) study of how clients and therpaists actually carry out meta-therapeutic discussions. It represents a sustained attempt to bridge the practice-research gap and highlights the conceptual and practical challenges in doing so. 42 audio-recorded pluralistic therapy sessions were sampled across seven therapist-client pairs. Before carrying out the CA study proper, it was necessary to conceptually link broad descriptions of meta-therapeutic dialogue to participants’ concrete actions in therapy sessions. This involved a review of related concepts (Chapter Two), as well as a detailed conceptualization of how therapists’ stocks of interactional knowledge (SIKs) (Peräkylä & Vehviläinen, 2003) regarding meta-therapeutic dialogue might demonstrably link with their concrete actions as described by CA findings (Chapters Three through Five). Therapists’ questions to clients about what might be helpful were selected as a likely site for meta-therapeutic dialogue and were subjected to an in-depth CA investigation of the practical issues participants themselves treated as important in their interactions around these questions (Chapters Six through Eight). Findings show how some apparent opportunities for meta-therapeutic dialogue are less facilitative of clients’ independent input, and can sometimes be interactionally coercive. There is evidence that facilitating dialogical opportunities for talking about what might be helpful may require the therapist to move back-and-forth between opposing positions, such as treating the client as potentially unknowing but still also holding open a space for their contribution. These findings extend existing SIKs regarding meta-therapeutic dialogue by specifying some concrete considerations therapists orient to during such endeavours. Some practical similarities between meta-therapeutic dialogue and problem-solving/solution-focused approaches are also highlighted.

Les glioblastomes comptent parmi les cas les plus répandus et agressifs de tumeurs cérébrales. Ils sont généralement traités avec une combinaison de résection chirurgicale, suivie de chimiothérapie et radiothérapie. Cependant, le caractère infiltrant de la tumeur rend son traitement particulièrement délicat. La personnalisation de modèles biophysiques permet d’automatiser la mise au point de thérapies spécifiques au patient, en maximisant les chances de survie. Dans cette thèse nous nous sommes attachés à élaborer des outils permettant de personnaliser la radiothérapie des glioblastomes. Nous avons tout d’abord étudié l’impact de la prise en compte de l’œdème vasogénique. Notre étude rétrospective se fonde sur une base de donnée de patients traités avec un médicament anti-angiogénique, révélant a posteriori la présence de l’œdème. Ensuite, nous avons étudié le lien entre l’incertitude due à la segmentation de la tumeur et la distribution de la dose. Pour se faire, nous avons mis au point une méthode permettant d’échantillonner efficacement de multiples segmentations réalistes, à partir d’une unique segmentation clinique. De plus, nous avons personnalisé un modèle de croissance tumorale aux images IRM de sept patients. La méthode Bayésienne adoptée permet notamment d’estimer l’incertitude sur les paramètres personnalisés. Finalement, nous avons montré comment cette personnalisation permet de définir automatiquement la dose à prescrire au patient, en combinant le modèle de croissance tumoral avec un modèle de réponse à la dose délivrée. Les résultats prometteurs présentés ouvrent de nouvelles perspectives pour la personnalisation de la radiothérapie des tumeurs cérébrales. / Glioblastomas are among the most common and aggressive primary brain tumors. It is usually treated with a combination of surgical resection, followed with concurrent chemo- and radiotherapy. However, the infiltrative nature of the tumor makes its control particularly challenging. Biophysical model personalization allows one to automatically define patient specific therapy plans which maximize survival rates. In this thesis, we focused on the elaboration of tools to personalize radiotherapy planning. First, we studied the impact of taking into account the vasogenic edema into the planning. We studied a database of patients treated with anti-angiogenic drug, revealing a posteriori the presence of the edema. Second, we studied the relationship between the uncertainty in the tumor segmentation and dose distribution. For that, we present an approach in order to efficiently sample multiple plausible segmentations from a single expert one. Third, we personalized a tumor growth model to seven patients’ MR images. We used a Bayesian approach in order to estimate the uncertainty in the personalized parameters of the model. Finally, we showed how combining a personalized model of tumor growth with a dose response model could be used to automatically define patient specific dose distribution. The promising results of our approaches offer new perspectives for personalized therapy planning.

FundaÃÃo Cearense de Apoio ao Desenvolvimento Cientifico e TecnolÃgico / O presente trabalho tem por objetivo discutir a personalizaÃÃo da polÃtica por parte dos representantes ao longo do mandato. Analisa-se como a utilizaÃÃo de media digitais tem participaÃÃo nesse fenÃmeno. Explora-se como agentes do campo acabam se apropriando dessas redes digitais para aumentar sua visibilidade, atravÃs da promoÃÃo da prÃpria imagem, a fim de influenciar favoravelmente a opiniÃo pÃblica. Esses meios sÃo frequentemente encarados com o potencial de aproximar a polÃtica dos cidadÃos, aumentando o engajamento da esfera civil. Examinando os perfis pessoais de dois deputados federais no Twitter â Marco Feliciano (PSC-SP) e Jean Wyllys (PSOL-RJ) â, busca-se estudar o uso ―personalizado‖ dos media digitais por estes agentes polÃticos. Para isso, foram coletados e estudados os tweets publicados nos meses de setembro, outubro e novembro de 2013, a fim de contemplar os perfis dos dois parlamentares de maneira mais contÃnua, sem se restringir a um acontecimento especÃfico, enfatizando-se os perfis dos diÃlogos travados entre os parlamentares e os usuÃrios. Para classificaÃÃo dos tweets, utilizou-se anÃlise de conteÃdo das mensagens na tentativa de dimensionar a gestÃo de imagem dos deputados a partir do que foi publicado em seus perfis. A anÃlise das mensagens revelou uma composiÃÃo de uma imagem pÃblica de dois deputados bastante integrados com as novas tecnologias e que, pelo menos na aparÃncia, estÃo dispostos a manter um diÃlogo direto com os cidadÃos. Ao gerenciarem suas imagens pÃblicas, tÃm-se principalmente os deputados falando sobre as posiÃÃes detidas e realizaÃÃes pessoais, ligando, por vezes, a vÃrias remodelaÃÃes e compromissos que ocorrem no seu cotidiano. Portanto, trata-se de uma informaÃÃo pessoal, mas de natureza profissional e nÃo oferece a seus seguidores um senso do processo polÃtico interno para a tomada de certas decisÃes. PorÃm, Ã inegÃvel que a utilizaÃÃo de ferramentas digitais aproxima representantes e representados e, por consequÃncia, tambÃm aproxima o discurso polÃtico da sociedade, ainda que as trocas sejam limitadas.