Spelling suggestions: "subject:"19kontext–aware"" "subject:"19kontext–zware""
251 |
Modelagem do domínio do processo de gerenciamento de nível de serviço do padrão ITIL : uma abordagem usando ontologias de fundamentação e sua aplicação na plataforma InfrawareCosta, André Cypriano Monteiro 15 August 2008 (has links)
Made available in DSpace on 2016-12-23T14:33:41Z (GMT). No. of bitstreams: 1
Andre Costa - Dissertacao - parte 1.pdf: 1053237 bytes, checksum: 8e90f2576db5227bb4eb33f34158733a (MD5)
Previous issue date: 2008-08-15 / This work presents a domain conceptual model of ITIL Service Level Management by using foundational ontologies, and also presents the application of the definied concepts in this model in the service provider module of Infraware context-aware service platform. The study of the ITIL service management process and conceptual modeling of this domain is an attempt to give the first steps in what refers to the formalization of this domain concepts and its use in context-aware service platforms, in particular. A case study is presented to validate the model and to verify its applicability in the generation of management information and of control to the service provider module of the platform and its customers / Este trabalho apresenta um modelo conceitual do domínio do processo de Gerenciamento de Nível de Serviço da biblioteca ITIL utilizando ontologias de fundamentação, e a aplicação dos conceitos definidos neste modelo no módulo provedor de serviço da plataforma de serviços sensíveis ao contexto Infraware. O estudo do processo de gerenciamento de serviços ITIL e a modelagem conceitual deste domínio visam dar os primeiros passos no que se refere à formalização dos conceitos deste domínio e a sua utilização em plataformas de serviços sensíveis ao contexto, em particular. Um estudo de caso é desenvolvido como forma de validar o modelo e de comprovar sua aplicabilidade na geração de informações de gerenciamento e de controle para o módulo provedor de serviço da plataforma e para seus clientes
|
252 |
ConBus: Uma Plataforma de Middleware de Integração de Sensores para o Desenvolvimento de Aplicações Móveis Sensíveis ao Contexto / ConBus: A Sensor Integration Middleware Platform for Mobile Context-Aware Application DevelopmentSÁ, Marcio Pereira de 26 April 2010 (has links)
Made available in DSpace on 2014-07-29T14:57:53Z (GMT). No. of bitstreams: 1
Marcio pereira.pdf: 5468645 bytes, checksum: c32705115b5d19bad90c4f72b43826ce (MD5)
Previous issue date: 2010-04-26 / In spite of the great evolution and dissemination of mobile devices and embedded
sensors, development of ubiquitous applications is still a complex task mainly due to the
great diversity of context information and the abundance of sensor technologies. In this
scenario, middleware systems are responsible mediating communication between contextaware
applications and sensors. This responsibility envolves many services such as sensor
communication protocols, asynchronous communication, context information reasoning.
In spite of their importance for mobile context-aware applications, the development of
middleware platforms for context provisioning is also a very complex task, specially in
terms of sensor module integration to these platforms. This happens due to many factors,
such as: i) huge complexity to develop sensor modules; ii) dificulties of reuse of sensor
modules; and iii) sensor module life cycle management. This work proposes a context
provisioning middleware architecture for mobile devices named ConBus (Context Bus)
that implements development, reuse, deployment and dynamic activation strategies for
sensor modules. / Apesar da grande evolução e disseminação dos dispositivos móveis e sensores acoplados,
desenvolver aplicações ubíquas ainda é uma tarefa complexa, principalmente, devido à
grande diversidade de informações contextuais e à abundância de tecnologias de sensoriamento.
Nesse cenário, sistemas de middleware assumem a responsabilidade de intermediar
a comunicação entre as aplicações sensíveis ao contexto e os sensores que são
as fontes de informações contextuais. Essa responsabilidade envolve diversos serviços,
como implementar protocolos de comunicação com sensores heterogêneos, disponibilizar
a comunicação assíncrona, possibilitar a inferência de informações contextuais, além da
manutenção de modelos de contexto de alto nível. Entretanto, o desenvolvimento de plataformas
de middleware para a provisão de contexto também é uma tarefa muito complexa,
especialmente com relação à integração de módulos de sensoriamento a tais infraestruturas.
Esses módulos de sensoriamento são os componentes de software das aplicações
responsáveis pelo acesso aos dados de contexto coletados pelos sensores. Dentre os principais
problemas relativos à essa integração estão: i) a complexidade inerente ao desenvolvimento
de módulos de sensoriamento, que usualmente envolvem chamadas de baixo
nível ao sistema operacional ou exigem a implementação de protocolos de comunicação
para acesso a sensores remotos; ii) dificuldade de reutilização dos módulos de sensoriamento
devido à falta de mecanismos que facilitem a disponibilização e a manutenção de
tais módulos; e iii) o gerenciamento do ciclo de vida de módulos de sensoriamento acoplados
à plataforma. Com o propósito de lidar com tais desafios, este trabalho propõe uma
arquitetura de middleware para provisão de contexto em dispositivos móveis, denominada
ConBus (Context Bus), que implementa estratégias de desenvolvimento, reutilização, implantação
e ativação dinâmica de módulos de sensoriamento, fazendo uso racional dos
recursos computacionais do dispositivo.
|
253 |
MHNCS: um middleware para o desenvolvimento de aplicações móveis cientes de contexto com requisitos de QoC / MHNCS: um middleware para o desenvolvimento de aplicações móveis cientes de contexto com requisitos de QoC / MNCS: a middleware for development of context-aware mobile applications with requirements of QoC / MNCS: a middleware for development of context-aware mobile applications with requirements of QoCPinheiro, Dejailson Nascimento 06 August 2014 (has links)
Made available in DSpace on 2016-08-17T14:53:29Z (GMT). No. of bitstreams: 1
DISSERTACAO Dejailson Nascimento Pinheiro.pdf: 1433962 bytes, checksum: 4173dad207f09fa2033a834f86a5d4b7 (MD5)
Previous issue date: 2014-08-06 / Mobile Social Networks (MSNs) are social structures in which members relate in groups and interaction is accomplished through information
and communication technologies using portable devices and wireless network technologies. Healthcare is one among the many possible areas of RSMs application.
The MobileHealthNet project, developed in partnership by UFMA and PUC-Rio, aims to develop a middleware that allows access to social networks and facilitate the
development of collaborative services targeting the health domain, the exchange of experiences and communication between patients and health professionals, as well
as a better management of health resources by government agencies. An important aspect in the development of the MobileHealthNet middleware is the infrastructure necessary for the gathering, distribution and processing of context data. In this master thesis we propose a software infrastructure incorporated to
the MobileHealthNet middleware that allows the specification, acquisition, validation and distribution of context data, considering quality requirements, making them available to context-aware applications. The distribution of context data is based on
a data-centric the publish/subscribe model, using the OMG-DDS specification. / Redes Sociais Móveis (RSMs) são estruturas sociais em que seus membros relacionam-se em grupos e a interação é realizada através de tecnologias de informação
e comunicação utilizando dispositivos portáteis com acesso a tecnologias de rede sem fio. Entre os muitos domínios de aplicação das RSMs, temos a área da saúde.
O projeto MobileHealthNet, desenvolvido em parceria pela UFMA e PUC-Rio, tem por objetivo desenvolver um middleware que permita o acesso às redes sociais e
facilite o desenvolvimento de serviços colaborativos para o setor da saúde, a troca de experiências e a comunicação entre pacientes e profissionais da saúde, além de uma melhor gestão dos recursos da saúde por órgãos governamentais. Um aspecto importante no desenvolvimento do middleware proposto pelo projeto MobileHealthNet é a infraestrutura necessária para a coleta, distribuição e processamento de dados de
contexto. Neste trabalho de mestrado é proposta uma infraestrutura de software incorporada ao middleware MobileHealthNet que permite a especificação, obtenção,
validação e distribuição de dados de contexto, considerando requisitos de qualidade, tornando-os disponíveis a aplicações sensíveis ao contexto. A distribuição dos dados de contexto é baseado no modelo publish/subscribe centrado em dados, utilizando-se a
especificação OMG-DDS.
|
254 |
UMA ARQUITETURA DE SENSORIAMENTO REMOTO SENSÍVEL AO CONTEXTO PARA IRRIGAÇÃO / A CONTEXT-AWARE REMOTE SENSING ARCHITECTURE FOR IRRIGATIONBoufleuer, Rafael 07 August 2015 (has links)
Considering that irrigation occupies the major percentage of consumed water in agriculture
and that Brazil intends to expand considerably its irrigated area in the coming decades, there
is a growing need in improving the irrigation water management, mostly in regions with limited
availability of water or occurrences of hydric deficiency periods. Therefore, the increase of
technologies for performing irrigation using environmental information is becoming important
because they enable the maximization of water and energy consumption, maintaining or even
improving the yield and quality of agricultural production. This work proposes a context-aware
remote sensing architecture for irrigation and its contributions are: (a) the development of two
prototypes of a moisture and precipitation meter based on the proposed architecture using open
hardware technologies; (b) the development of a context taxonomy that defines the types of
information that can be used in the prototype s architecture; and (c) a comparison between two
different types of soil moisture sensors. For the validation of the architecture, two case studies
were realized to verify the correct functioning of the architecture components, as well as a data
collection was performed to make the comparison between a resistive low-cost sensor produced
at the Federal University of Santa Maria, and a high-accuracy and high-cost frequency domain
reflectometry (FDR) sensor (CS616 - Campbell Scientific, United States). The results obtained
from the analyzed data were satisfactory, where it was verified that the architecture is viable,
meeting the requirements to which it has proposed. In addition, the comparison performed
showed a determination coefficient of up to 95 % between the low-cost resistive soil moisture
sensors and the soil moisture sensors CS616 of Campbell
R . / Considerando que a irrigação ocupa a maior porcentagem da água consumida na agricultura,
e que o Brasil pretende expandir consideravelmente a sua área irrigada nas próximas
décadas, há uma crescente necessidade de melhorar o manejo da água de irrigação, principalmente
em regiões com reduzida disponibilidade de água ou ocorrência de períodos de deficiência
hídrica. Por isso, o incremento de tecnologias na realização da irrigação utilizando-se de
informações referentes ao ambiente, estão tornando-se importantes por permitirem a maximização
do uso da água e energia, mantendo, ou mesmo melhorando, o rendimento e a qualidade
da produção agrícola. Este trabalho propõe uma arquitetura de sensoriamento remoto sensível
ao contexto para irrigação, e suas contribuições são: (a) o desenvolvimento de dois protótipos
de um medidor de umidade e pluviometria baseados na arquitetura proposta utilizando tecnologias
open hardware; (b) o desenvolvimento de uma taxonomia de contextos que define os tipos
de informações que podem ser utilizadas na arquitetura dos protótipos; e (c) uma comparação
entre dois tipos distintos de sensores de umidade do solo. Para a validação da arquitetura, foram
realizados dois estudos de caso para verificar o correto funcionamento dos componentes da
arquitetura, bem como a coleta de dados para efetuar a comparação entre um sensor resistivo
de baixo custo produzido na Universidade Federal de Santa Maria, e um sensor de alta precisão
e custo que utiliza a técnica de reflectometria no domínio da frequência (FDR) (CS616 -
Campbell Scientific, Estados Unidos). Os resultados obtidos com os dados analisados foram satisfatórios,
onde verificou-se que a arquitetura atende aos requisitos aos quais se propôs. Além
disso, a comparação realizada apresentou coeficientes de determinação de até 95% entre os sensores
resistivos de umidade do solo de baixo custo e os sensores de umidade do solo CS616 da
Campbell
R .
|
255 |
Development of an energy efficient, robust and modular multicore wireless sensor network / Développement d’un capteur multicoeur sans fil à énergie efficient, robuste et modulaireShi, Hong-Ling 23 January 2014 (has links)
Le réseau de capteurs sans fil est une technologie clé du 21ème siècle car ses applications sont nombreuses et diverses. Cependant le réseau de capteurs sans fil est un système à très forte contrainte de ressources. En conséquence, les techniques utilisées pour le développement des systèmes embarqués classiques ne peuvent être appliquées. Aujourd’hui les capteurs sans fil ont été réalisés en utilisant une architecture monoprocesseur. Cette approche ne permet pas de réaliser un capteur sans fil robuste et à énergie efficiente pour les applications telles que agriculture de précision (en extérieur) et télémédecine. Les travaux menés dans le cadre de cette thèse ont pour but de développer une nouvelle approche pour la réalisation d’un capteur sans fil en utilisant une architecture multicoeur pour permettre à la fois d’augmenter sa robustesse et sa durée de vie (minimiser sa consommation énergétique). / The wireless sensor network is a key technology in the 21st century because it has multitude applications and it becomes the new way of interaction between physical environment and computer system. Moreover, the wireless sensor network is a high resource constraint system. Consequently, the techniques used for the development of traditional embedded systems cannot be directly applied. Today wireless sensor nodes were implemented by using only one single processor architecture. This approach does not achieve a robust and efficient energy wireless sensor network for applications such as precision agriculture (outdoor) and telemedicine. The aim of this thesis is to develop a new approach for the realization of a wireless sensor network node using multicore architecture to enable to increase both its robustness and lifetime (reduce energy consumption).
|
256 |
Uma abordagem ciente de contexto e embasada por feedbacks para o gerenciamento de handovers em ambientes NGN / A context-aware and feedback-based approach for handover management in NGNRoberto Rigolin Ferreira Lopes 20 June 2012 (has links)
A evolução da computação móvel melhora a capacidade de comunicação e colaboração das pessoas. Os principais pilares desta transformação são: o desenvolvimento e produção de dispositivos móveis com capacidade multimídia e equipados com duas ou mais interfaces de rede, a disponibilidade de conectividade sem fio ubíqua e a popularização de aplicações sociais online. As redes sociais online merecem destaque pelas funcionalidades que permitem a criação e compartilhamento de conteúdo digital dentro de círculos sociais, também chamado de mídia social. Serviços na web anexam a localização geográfica do dispositivo ao conteúdo digital, criando as chamadas mídias sociais baseadas em localização. Equipadas com seus telefones e tablets, as pessoas estão criando e consumindo mídias sociais em qualquer lugar. Entretanto, é um desafio manter tais dispositivos móveis conectados nos ambientes de rede sem fio atuais e de próxima geração e.g., múltiplos provedores de acesso e múltiplas tecnologias de comunicação. Pesquisas recentes propõem componentes para o gerenciamento de conectividade sem fio que fazem uso simultâneo do contexto de conectividade atual e de um conjunto destes dados coletados no passado. Tais componentes são preditores de mobilidade, mecanismos de handover ou gerenciadores de mobilidade que utilizam dados de contexto de conectividade de forma particular para atingir seus propósitos. Na presente investigação, propomos uma metodologia que orquestra os principais componentes de gerenciamento de conectividade em um laço retro alimentado. Argumentamos que a coleta de dados de contexto de conectividade pode ser projetada como um sistema de sensoreamento, cujo sensores são as interfaces de rede sem fio. Como parte deste sistema de sensoriamento, os círculos sociais podem assistir o gerenciamento de conectividade compartilhando dados de contexto de conectividade. A ideia central é utilizar serviços baseados em localização para compartilhar dados de contexto de conectividade dentro dos círculos sociais. Desta forma, as redes sociais online adicionam escala para o sistema e permite colaboração em volta de dados de contexto recentes, locais, personalizados e sociais. O objetivo é melhorar experiências de conectividade sem fio e.g., métricas de QoS (Quality of Service) como: vazão, latência e qualidade do sinal. Relatamos como os dados de contexto de conectividade são manipulados com um modelo baseado em grafos e métricas como: intensidade do vértice e grau centralidade. Com isso, identificamos áreas com alta densidade de handovers, definimos a reputação dos usuários e revelamos a cobertura das redes. Resultados de experimentos mostram que a colaboração pode melhorar métricas de QoS de ~18 a ~30% se comparado ao uso de um preditor de mobilidade ou um sistema operacional moderno, respectivamente. Esta discussão se desdobra com foco na viabilidade da solução em termos de sobrecarga de armazenamento e consumo de energia. Os promissores resultados experimentais indicam que nossa solução pode melhorar experiências de conectividade sem fio de usuários móveis / The evolution of mobile computing improves communication and collaboration among people. The main pillars of this transformation are: the development and production of mobile devices with multimedia capabilities and equipped with two or more network interfaces, the availability of ubiquitous wireless connectivity and the popularity of online social applications. Online social networks noteworthy features that allow for the creation and sharing of digital content within social circles, also called textit Social Media. Web Services attach the geographic location of the device to the digital content, creating the so-called textit location-based social media. Equipped with their phones and tablets, people are creating and consuming social media anywhere. However, it is a challenge to keep such mobile devices connected in current and next generation wireless network environments textit e.g., multiple ISPs (Internet Service Provider) and multiple communication technologies. Recent researches proposes components for managing wireless connectivity that make simultaneous use of the current and past connectivity context data. Such components are mobility predictors, handovers mechanisms or mobility managers that use connectivity context data in a particular way to achieve its purposes. In this research, we propose feasiable a methodology that orchestrates the main components of the connectivity management in a feedback loop. We argue that the process of gathering connectivity context data can be designed as a sensing system, whose sensors are wireless network interfaces. As part of this sensing system, the social circles may assist the management of connectivity by sharing connectivity context data. The main idea is to use location-based services to share connectivity context data within social circles. Thus, online social networks add scale to the system and enables collaboration around recent, local, and social context data. The goal is to enhance wireless connectivity experiences in terms of QoS ( textit Quality of Service) metrics textit e.g., throughput, latency and signal quality. We report how this data is handled using complex networks metrics e.g., vertexs strength and centrality degree, to identify high density handover areas, define the mobile users reputation and to reveal the networks coverage. Real experiments showed that collaboration can improve QoS metrics from ~18 to ~30% if compared to just use a mobility predictor or a modern operational system, respectively. The discussion unfolds with focus on the collaborations efficiency as function of time, number of users, discovered area size and mobility patterns. The promising experimental results indicate that our solution can enhance mobile users wireless connectivity experiences
|
257 |
Context-aware ranking : from search to dialogueZhu, Yutao 03 1900 (has links)
Les systèmes de recherche d'information (RI) ou moteurs de recherche ont été largement utilisés pour trouver rapidement les informations pour les utilisateurs. Le classement est la fonction centrale de la RI, qui vise à ordonner les documents candidats dans une liste classée en fonction de leur pertinence par rapport à une requête de l'utilisateur. Alors que IR n'a considéré qu'une seule requête au début, les systèmes plus récents prennent en compte les informations de contexte. Par exemple, dans une session de recherche, le contexte de recherche tel que le requêtes et interactions précédentes avec l'utilisateur, est largement utilisé pour comprendre l'intention de la recherche de l'utilisateur et pour aider au classement des documents. En plus de la recherche ad-hoc traditionnelle, la RI a été étendue aux systèmes de dialogue (c'est-à-dire, le dialogue basé sur la recherche, par exemple, XiaoIce), où on suppose avoir un grand référentiel de dialogues et le but est de trouver la réponse pertinente à l'énoncé courant d'un utilisateur. Encore une fois, le contexte du dialogue est un élément clé pour déterminer la pertinence d'une réponse.
L'utilisation des informations contextuelles a fait l'objet de nombreuses études, allant de l'extraction de mots-clés importants du contexte pour étendre la requête ou l'énoncé courant de dialogue, à la construction d'une représentation neuronale du contexte qui sera utilisée avec la requête ou l'énoncé de dialogue pour la recherche. Nous remarquons deux d'importantes insuffisances dans la littérature existante. (1) Pour apprendre à utiliser les informations contextuelles, on doit extraire des échantillons positifs et négatifs pour l'entraînement. On a généralement supposé qu'un échantillon positif est formé lorsqu'un utilisateur interagit avec (clique sur) un document dans un contexte, et un un échantillon négatif est formé lorsqu'aucune interaction n'est observée. En réalité, les interactions des utilisateurs sont éparses et bruitées, ce qui rend l'hypothèse ci-dessus irréaliste. Il est donc important de construire des exemples d'entraînement d'une manière plus appropriée. (2) Dans les systèmes de dialogue, en particulier les systèmes de bavardage (chitchat), on cherche à trouver ou générer les réponses sans faire référence à des connaissances externes, ce qui peut facilement provoquer des réponses non pertinentes ou des hallucinations. Une solution consiste à fonder le dialogue sur des documents ou graphe de connaissances externes, où les documents ou les graphes de connaissances peuvent être considérés comme de nouveaux types de contexte. Le dialogue fondé sur les documents et les connaissances a été largement étudié, mais les approches restent simplistes dans la mesure où le contenu du document ou les connaissances sont généralement
concaténés à l'énoncé courant. En réalité, seules certaines parties du document ou du
graphe de connaissances sont pertinentes, ce qui justifie un modèle spécifique pour leur sélection.
Dans cette thèse, nous étudions le problème du classement de textes en tenant compte du contexte dans le cadre de RI ad-hoc et de dialogue basé sur la recherche. Nous nous concentrons sur les deux problèmes mentionnés ci-dessus. Spécifiquement, nous proposons des approches pour apprendre un modèle de classement pour la RI ad-hoc basée sur des exemples d'entraîenemt sélectionnés à partir d'interactions utilisateur bruitées (c'est-à-dire des logs de requêtes) et des approches à exploiter des connaissances externes pour la recherche de réponse pour le dialogue. La thèse est basée sur cinq articles publiés. Les deux premiers articles portent sur le classement contextuel des documents. Ils traitent le problème ovservé dans les études existantes, qui considèrent tous les clics dans les logs de recherche comme des échantillons positifs, et prélever des documents non cliqués comme échantillons négatifs. Dans ces deux articles, nous proposons d'abord une stratégie d'augmentation de données non supervisée pour simuler les variations potentielles du comportement de l'utilisateur pour tenir compte de la sparcité des comportements des utilisateurs. Ensuite, nous appliquons l'apprentissage contrastif pour identifier ces variations et à générer une représentation plus robuste du comportement de l'utilisateur. D'un autre côté, comprendre l'intention de recherche dans une session de recherche peut représentent différents niveaux de difficulté - certaines intentions sont faciles à comprendre tandis que d'autres sont plus difficiles et nuancées. Mélanger directement ces sessions dans le même batch d'entraînement perturbera l'optimisation du modèle. Par conséquent, nous proposons un cadre d'apprentissage par curriculum avec des examples allant de plus faciles à plus difficiles. Les deux méthodes proposées obtiennent de meilleurs résultats que les méthodes existantes sur deux jeux de données de logs de requêtes réels.
Les trois derniers articles se concentrent sur les systèmes de dialogue fondé les documents/connaissances. Nous proposons d'abord un mécanisme de sélection de contenu pour le dialogue fondé sur des documents. Les expérimentations confirment que la sélection de contenu de document pertinent en fonction du contexte du dialogue peut réduire le bruit dans le document et ainsi améliorer la qualité du dialogue. Deuxièmement, nous explorons une nouvelle tâche de dialogue qui vise à générer des dialogues selon une description narrative. Nous avons collecté un nouveau jeu de données dans le domaine du cinéma pour nos expérimentations. Les connaissances sont définies par une narration qui décrit une partie du scénario du film (similaire aux dialogues). Le but est de créer des dialogues correspondant à la narration. À cette fin, nous concevons un nouveau modèle qui tient l'état de la couverture de la narration le long des dialogues et déterminer la partie non couverte pour le prochain tour. Troisièmement, nous explorons un modèle de dialogue proactif qui peut diriger de manière proactive le dialogue dans une direction pour couvrir les sujets requis. Nous concevons un module de prédiction explicite des connaissances pour sélectionner les connaissances pertinentes à utiliser. Pour entraîner le processus de sélection, nous générons des signaux de supervision par une méthode heuristique. Les trois articles examinent comment divers types de connaissances peuvent être intégrés dans le dialogue.
Le contexte est un élément important dans la RI ad-hoc et le dialogue, mais nous soutenons que le contexte doit être compris au sens large. Dans cette thèse, nous incluons à la fois les interactions précédentes avec l'utilisateur, le document et les connaissances dans le contexte. Cette série d'études est un pas dans la direction de l'intégration d'informations contextuelles diverses dans la RI et le dialogue. / Information retrieval (IR) or search systems have been widely used to quickly find desired information for users. Ranking is the central function of IR, which aims at ordering the candidate documents in a ranked list according to their relevance to a user query. While IR only considered a single query in the early stages, more recent systems take into account the context information. For example, in a search session, the search context, such as the previous queries and interactions with the user, is widely used to understand the user's search intent and to help document ranking. In addition to the traditional ad-hoc search, IR has been extended to dialogue systems (i.e., retrieval-based dialogue, e.g., XiaoIce), where one assumes a large repository of previous dialogues and the goal is to retrieve the most relevant response to a user's current utterance. Again, the dialogue context is a key element for determining the relevance of a response.
The utilization of context information has been investigated in many studies, which range from extracting important keywords from the context to expand the query or current utterance, to building a neural context representation used with the query or current utterance for search. We notice two important insufficiencies in the existing literature. (1) To learn to use context information, one has to extract positive and negative samples for training. It has been generally assumed that a positive sample is formed when a user interacts with a document in a context, and a negative sample is formed when no interaction is observed. In reality, user interactions are scarce and noisy, making the above assumption unrealistic. It is thus important to build more appropriate training examples. (2) In dialogue systems, especially chitchat systems, responses are typically retrieved or generated without referring to external knowledge. This may easily lead to hallucinations. A solution is to ground dialogue on external documents or knowledge graphs, where the grounding document or knowledge can be seen as new types of context. Document- and knowledge-grounded dialogue have been extensively studied, but the approaches remain simplistic in that the document content or knowledge is typically concatenated to the current utterance. In reality, only parts of the grounding document or knowledge are relevant, which warrant a specific model for their selection.
In this thesis, we study the problem of context-aware ranking for ad-hoc document ranking and retrieval-based dialogue. We focus on the two problems mentioned above. Specifically, we propose approaches to learning a ranking model for ad-hoc retrieval based on training examples selected from noisy user interactions (i.e., query logs), and approaches to exploit external knowledge for response retrieval in retrieval-based dialogue. The thesis is based on five published articles.
The first two articles are about context-aware document ranking. They deal with the problem in the existing studies that consider all clicks in the search logs as positive samples, and sample unclicked documents as negative samples. In the first paper, we propose an unsupervised data augmentation strategy to simulate potential variations of user behavior sequences to take into account the scarcity of user behaviors. Then, we apply contrastive learning to identify these variations and generate a more robust representation for user behavior sequences. On the other hand, understanding the search intent of search sessions may represent different levels of difficulty -- some are easy to understand while others are more difficult. Directly mixing these search sessions in the same training batch will disturb the model optimization. Therefore, in the second paper, we propose a curriculum learning framework to learn the training samples in an easy-to-hard manner. Both proposed methods achieve better performance than the existing methods on two real search log datasets.
The latter three articles focus on knowledge-grounded retrieval-based dialogue systems. We first propose a content selection mechanism for document-grounded dialogue and demonstrate that selecting relevant document content based on dialogue context can effectively reduce the noise in the document and increase dialogue quality. Second, we explore a new task of dialogue, which is required to generate dialogue according to a narrative description. We collect a new dataset in the movie domain to support our study. The knowledge is defined as a narrative that describes a part of a movie script (similar to dialogues). The goal is to create dialogues corresponding to the narrative. To this end, we design a new model that can track the coverage of the narrative along the dialogues and determine the uncovered part for the next turn. Third, we explore a proactive dialogue model that can proactively lead the dialogue to cover the required topics. We design an explicit knowledge prediction module to select relevant pieces of knowledge to use. To train the selection process, we generate weak-supervision signals using a heuristic method. All of the three papers investigate how various types of knowledge can be integrated into dialogue.
Context is an important element in ad-hoc search and dialogue, but we argue that context should be understood in a broad sense. In this thesis, we include both previous interactions and the grounding document and knowledge as part of the context. This series of studies is one step in the direction of incorporating broad context information into search and dialogue.
|
258 |
Un système de télésanté contextuel avec support de qualité de service pour le maintien à domicile / A context-aware and QoS-aware telehomecare systemNourizadeh, Shahram 06 July 2011 (has links)
Cette thèse est une thèse CIFRE entre le LORIA et la société MEDETIC et porte sur la conception des systèmes de télésurveillance pour le maintien à domicile des personnes âgées.Le système est conçu aux personnes âgées qui veulent passer leur vieillesse dans leur propre maison, à cause de son potentiel pour augmenter l'indépendance et la qualité de la vie. Cela profiterait non seulement aux personnes âgées qui veulent vivre dans leur propre maison, mais aussi le système de santé publique en coupant des prix de façon significative.Mis à part la conception d'une architecture de télésurveillance « Vill’Âge® »basée sur des réseaux de capteurs hétérogènes (Domotique, IEEE802.15.4/Zigbee, Wifi, Bluetooth), la thèse a contribué essentiellement sur la proposition d'un protocole de clustering et de routage dans le réseau de capteurs sans fil avec une approche de la logique floue, et d'un middleware pour la collecte et le traitement des données des capteurs avec la gestion de la qualité de service comme particularité.Une première plateforme de test à été développée à Colmar (MEDETIC) et une seconde, plus complète et fait suite de cette thèse, est en cours de développement au LORIA (http://infositu.loria.fr/).Nous avons participé dans le concours de ministère de l'Enseignement supérieur et de la recherche avec un projet intitulé MPIGate: « Multi Protocol Interface and Gateway for telecare, environment Monitoring and Control » et nous avons gagné le prix de ce concours au niveau d’émergence. / This thesis is a thesis CIFRE between LORIA and the MEDETIC Company and focuses on the design of telehomecare system for the elderly.In addition to the design of a remote surveillance architecture “Vill’Âge ®” based on networks of heterogeneous sensor (home automation, IEEE802.15.4/Zigbee, Wifi, Bluetooth), thesis has contributed essentially on the proposal of a clustering and routing protocol in the network of wireless sensors with an approach to fuzzy logic, and of a middleware for the collection and processing of data from sensors with the management of the quality of service as a special feature.A first platform was developed at Colmar (MEDETIC) and a second who is more complete is under development at LORIA (http://infositu.loria.fr/).By using this system, MEDeTIC, offers a new concept of smart homes for the senior citizens, named in French “Maisons Vill’Âge”. The first housing schemes are being built in 2 departments of France. A flat is entirely equipped to act as a demonstrator and as laboratory of research and development.The system is designed for the elderly who wish to spend their old age in their own home, because of its potential to increase independence and quality of life. This would not only benefit the elderly who want to live in their own home, but also the national health care system by cutting costs significantly. Based on this PhD thesis, MPIGate, a “Multiprotocol Interface and Gateway for for telecare, environment”, has been developed. MPIGate was awarded in the competition of the Ministry of Higher Education and Research and OSEO 2010
|
259 |
DRARS, a dynamic risk-aware recommender system / DRARS, un système de recommandation dynamique sensible au risqueBouneffouf, Djallel 19 December 2013 (has links)
L’immense quantité d'information générée et gérée au quotidien par les systèmes d'information et leurs utilisateurs conduit inéluctablement à la problématique de surcharge d'information. Dans ce contexte, les systèmes de recommandation traditionnels fournissent des informations pertinentes aux utilisateurs. Néanmoins, avec la propagation récente des dispositifs mobiles (smartphones et tablettes), nous constatons une migration progressive des utilisateurs vers la manipulation d'environnements pervasifs. Le problème avec les approches de recommandation traditionnelles est qu'elles n'utilisent pas toute l'information disponible pour produire des recommandations. Davantage d’informations contextuelles pourraient être utilisées dans le processus de recommandation pour aboutir à des recommandations plus précises. Les systèmes de recommandation sensibles au contexte (CARS) combinent les caractéristiques des systèmes sensibles au contexte et des systèmes de recommandation afin de fournir des informations personnalisées aux utilisateurs dans des environnements ubiquitaires. Dans cette perspective où tout ce qui concerne l'utilisateur est dynamique, les contenus qu’il manipule et son environnement, deux questions principales doivent être adressées : i) Comment prendre en compte l'évolution des contenus de l’utilisateur? et ii) Comment éviter d’être intrusif, en particulier dans des situations critiques? En réponse à ces questions, nous avons développé un système de recommandation dynamique et sensible au risque appelé DRARS (Dynamic Risk-Aware Recommender System), qui modélise la recommandation sensible au contexte comme un problème de bandit. Ce système combine une technique de filtrage basée sur le contenu et un algorithme de bandit contextuel. Nous avons montré que DRARS améliore la stratégie de l'algorithme UCB (Upper Confidence Bound), le meilleur algorithme actuellement disponible, en calculant la valeur d'exploration la plus optimale pour maintenir un bon compromis entre exploration et exploitation basé sur le niveau de risque de la situation courante de l'utilisateur. Nous avons mené des expériences dans un contexte industriel avec des données réelles et des utilisateurs réels et nous avons montré que la prise en compte du niveau de risque de la situation de l'utilisateur augmentait significativement la performance du système de recommandation / The vast amount of information generated and maintained everyday by information systems and their users leads to the increasingly important concern of overload information. In this context, traditional recommender systems provide relevant information to the users. Nevertheless, with the recent dissemination of mobile devices (smartphones and tablets), there is a gradual user migration to the use of pervasive computing environments. The problem with the traditional recommendation approaches is that they do not utilize all available information for producing recommendations. More contextual parameters could be used in the recommendation process to result in more accurate recommendations. Context-Aware Recommender Systems (CARS) combine characteristics from context-aware systems and recommender systems in order to provide personalized recommendations to users in ubiquitous environments. In this perspective where everything about the user is dynamic, his/her content and his/her environment, two main issues have to be addressed: i) How to consider content evolution? and ii) How to avoid disturbing the user in risky situations?. In response to these problems, we have developed a dynamic risk sensitive recommendation system called DRARS (Dynamic Risk-Aware Recommender System), which model the context-aware recommendation as a bandit problem. This system combines a content-based technique and a contextual bandit algorithm. We have shown that DRARS improves the Upper Confidence Bound (UCB) policy, the currently available best algorithm, by calculating the most optimal exploration value to maintain a trade-off between exploration and exploitation based on the risk level of the current user's situation. We conducted experiments in an industrial context with real data and real users and we have shown that taking into account the risk level of users' situations significantly increases the performance of the recommender system
|
260 |
Intelligent flood adaptative contex-aware system / Système sensible et adaptatif au contexte pour la gestion intelligente de cruesSun, Jie 23 October 2017 (has links)
A l’avenir, l'agriculture et l'environnement vont pouvoir bénéficier de plus en plus de données hétérogènes collectées par des réseaux de capteurs sans fil (RCSF). Ces données alimentent généralement des outils d’aide à la décision (OAD). Dans cette thèse, nous nous intéressons spécifiquement aux systèmes sensibles et adaptatifs au contexte basés sur un RCSF et un OAD, dédiés au suivi de phénomènes naturels. Nous proposons ainsi une formalisation pour la conception et la mise en œuvre de ces systèmes. Le contexte considéré se compose de données issues du phénomène étudié mais également des capteurs sans fil (leur niveau d’énergie par exemple). Par l’utilisation des ontologies et de techniques de raisonnement, nous visons à maintenir le niveau de qualité de service (QdS) des données collectées (en accord avec le phénomène étudié) tant en préservant le fonctionnement du RCSF. Pour illustrer notre proposition, un cas d'utilisation complexe, l'étude des inondations dans un bassin hydrographique, est considéré. Cette thèse a produit un logiciel de simulation de ces systèmes qui intègre un système de simulation multi-agents (JADE) avec un moteur d’inférence à base de règles (Jess). / In the future, agriculture and environment will rely on more and more heterogeneous data collected by wireless sensor networks (WSN). These data are generally used in decision support systems (DSS). In this dissertation, we focus on adaptive context-aware systems based on WSN and DSS, dedicated to the monitoring of natural phenomena. Thus, a formalization for the design and the deployment of these kinds of systems is proposed. The considered context is established using the data from the studied phenomenon but also from the wireless sensors (e.g., their energy level). By the use of ontologies and reasoning techniques, we aim to maintain the required quality of service (QoS) level of the collected data (according to the studied phenomenon) while preserving the resources of the WSN. To illustrate our proposal, a complex use case, the study of floods in a watershed, is described. During this PhD thesis, a simulator for context-aware systems which integrates a multi-agent system (JADE) and a rule engine (Jess) has been developed.Keywords: ontologies, rule-based inferences, formalization, heterogeneous data, sensors data streams integration, WSN, limited resources, DSS, adaptive context-aware systems, QoS, agriculture, environment.
|
Page generated in 1.8048 seconds