A população idosa mundial está crescendo e, com ela, o número de diagnósticos de doenças relacionadas à velhice como, por exemplo, declínios cognitivos também. Essas doenças costumam afetar a autonomia do idoso no seu lar, especialmente no que se refere à realização de atividades diárias. Com isso em vista, é preciso empregar cuidadores e serviços de saúde que acabam por implicar em altos custos. Nesse sentido, surge a necessidade de conceber sistemas robustos, automatizados, utilizáveis e de baixo custo para a assistência pessoal. A concepção desses sistemas faz menção à área de Ambientes de Vivência Assistida. Portanto, esta dissertação propõe uma abordagem que beneficia os sistemas para Ambientes de Vivência Assistida com a capacidade de prever ações humanas para a facilitação de atividades diárias, particularmente quando declínios cognitivos relacionados à elas ocorrerem. Nesse sentido, foi concebido um meta-modelo semântico para a geração de modelos conceituais de contexto e de comportamento, compostos pelas ações humanas. A partir disso, a previsão de ações (informação de suporte) é realizada por um mecanismo de predição e inferência composto por um modelo semântico probabilístico. A abordagem é demonstrada através de um estudo de caso cujo cenário representa uma situação de declínio cognitivo, enfrentada por um usuário, que impede a condução de uma atividade diária. Então, o mecanismo de predição e inferência, utilizando o modelo semântico probabilístico, prevê qual a ação mais adequada que facilite a conclusão da atividade. Essa previsão é avaliada para aferir o quão bem um usuário seria auxiliado, isto é, se a operação prevista foi por ele realizada. Para isso, foi utilizado um dataset relacionado ao cenário do estudo de caso e medidas de desempenho como a precisão, a revocação e a medida-F. Os resultados dessa avaliação se mostraram promissores sendo, em média, 69,5% para a precisão, 100% para a revocação e 81% para a medida-F. As principais contribuições deste trabalho dizem respeito ao meta-modelo semântico a partir do qual pesquisas na área deste trabalho podem utilizar para gerar modelos de comportamento, e ao modelo semântico probabilístico que realiza predição através de raciocínio incerto sobre os modelos de comportamento, propiciando decisões mais precisas para auxiliar usuários com declínio cognitivo. / The world’s elderly population is growing and, with it, the number of diagnoses of diseases related to old age, such as cognitive declines as well. These diseases usually affect the autonomy of the elderly in their home, especially when it comes to performing daily activities. With this in mind, it is necessary to employ caregivers and health services that end up implying high costs. In this sense, the need arises to design robust, automated, usable and low-cost systems for personal assistance. The design of these systems makes reference to the area of Ambient Assisted Living. Therefore, this dissertation proposes an approach that benefits the Ambient Assisted Living systems with the ability to predict human actions for the facilitation of daily activities, particularly when cognitive declines related to them occur. In this sense, a semantic meta-model was conceived for the generation of conceptual models of context and behavior, composed by human actions. From this, the prediction of actions (information of support) is realized by a mechanism of prediction and inference composed by a probabilistic semantic model. The approach is demonstrated through a case study whose scenario represents a situation of cognitive decline, faced by a user, that prevents the conduct of a daily activity. Then, the prediction and inference mechanism, using the probabilistic semantic model, predicts the most appropriate action that facilitates the conclusion of the activity. This forecast is evaluated to gauge how well a user would be assisted, that is, if the intended operation was performed by him. For this, a dataset related to the case study scenario and performance measures such as precision, recall, and F-measure were used. The results of this evaluation are promising, averaging 69.5% for precision, 100% for recall and 81% for F-measure. The main contributions of this work are related to the semantic meta-model from which research in the area of this work can be used to generate behavioral models, and to the probabilistic semantic model that performs prediction through uncertain reasoning over behavior models, providing better decisions to help users with cognitive decline.

En dépit de leur caractère distribué, les chaînes logistiques peuvent se révéler très performantes dans les conditions idéales de production et d’échange. Toutefois, leur complexité les rend de plus en plus fragiles. Cette thèse propose des modèles et des méthodes pour l’analyse des risques, de façon à renforcer la robustesse et la résilience des CLs. Nous avons analysé ce domaine suivant une démarche ontologique à l’aide de la méthode KOD pour tirer les caractéristiques essentielles des CLs. En nous appuyant sur un état de l’art du domaine des risques dans les chaînes logistiques, et sur les bases de cas réels, nous avons identifié les indicateurs des vulnérabilités les plus significatifs. A partir des connaissances extraites, et des modèles mathématiques proposés dans la littérature, nous avons construit un modèle de CL multi-étages à l’aide de modèles ARIMA intégrant l’aspect aléatoire de la demande. Pour adapter ce modèle aux situations de vulnérabilité et de risques, nous avons ajouté des contraintes de capacité et de positivité sur les commandes et sur les stocks. Sous l’effet d’événements dangereux, certaines contraintes du système peuvent être atteintes et par conséquence, son évolution peut s’écarter fortement de la dynamique nominale. Nous avons proposé des indicateurs de vulnérabilités comme des indicateurs de fréquence des retards de livraison, ou de surcoût d’immobilisation de produits. Enfin, l’occurrence d’événements dangereux a été représentée par des scénarios. Nous avons alors obtenu des résultats de simulation sous MATLAB, qui nous ont permis d’évaluer leurs conséquences pour différentes configurations du système. / Despite their distributed nature, these supply chains can be very efficient in the ideal conditions of production and exchange. However, their complexity makes them more fragile. This dissertation proposes models and methods for risk analysis to enhance the robustness and resilience of SCs. We analyzed this area following an ontological approach using the KOD method. Based on state of the art in the field of risk in SCs, and on real cases, we identified the indicators of the most significant vulnerabilities. From the extracted knowledge and mathematical models proposed in the literature, we built the model of a multi-stage SC using ARIMA models incorporating the randomness of the demand. In order to adapt this model to situations of vulnerability and risk, we have added capacity and positivity constraints on orders and inventories. Under the impact of hazardous events or strong disturbances, some constraints of the system can be reached and therefore, its evolution may deviate considerably from the nominal dynamics or even become unstable. We proposed vulnerability indicators such as indicators of the frequency of delivery delays or costs due to the immobilization of products. Finally, scenarios were used to represent the occurrence of dangerous events. We then got simulation results in MATLAB, which allowed us to assess their consequences for different configurations of the system, especially for strong disturbances of information flows and physical flows .

Designing and discovering novel materials is challenging problem in many domains such as fuel additives, composites, pharmaceuticals, and so on. At the core of all this are models that capture how the different domain-specific data, information, and knowledge regarding the structures and properties of the materials are related to one another. This dissertation explores the difficult task of developing an artificial intelligence-based knowledge modeling environment, called Hybrid Ontology-Learning Materials Engineering System (HOLMES) that can assist humans in populating a materials science and engineering ontology through automatic information extraction from journal article abstracts. While what we propose may be adapted for a generic materials engineering application, our focus in this thesis is on the needs of the pharmaceutical industry. We develop the Columbia Ontology for Pharmaceutical Engineering (COPE), which is a modification of the Purdue Ontology for Pharmaceutical Engineering. COPE serves as the basis for HOLMES. The HOLMES framework starts with journal articles that are in the Portable Document Format (PDF) and ends with the assignment of the entries in the journal articles into ontologies. While this might seem to be a simple task of information extraction, to fully extract the information such that the ontology is filled as completely and correctly as possible is not easy when considering a fully developed ontology. In the development of the information extraction tasks, we note that there are new problems that have not arisen in previous information extraction work in the literature. The first is the necessity to extract auxiliary information in the form of concepts such as actions, ideas, problem specifications, properties, etc. The second problem is in the existence of multiple labels for a single token due to the existence of the aforementioned concepts. These two problems are the focus of this dissertation. In this work, the HOLMES framework is presented as a whole, describing our successful progress as well as unsolved problems, which might help future research on this topic. The ontology is then presented to help in the identification of the relevant information that needs to be retrieved. The annotations are next developed to create the data sets necessary for the machine learning algorithms to perform. Then, the current level of information extraction for these concepts is explored and expanded. This is done through the introduction of entity feature sets that are based on previously extracted entities from the entity recognition task. And finally, the new task of handling multiple labels for tagging a single entity is also explored by the use of multiple-label algorithms used primarily in image processing.

This project investigates Saharan pastoralist rock-art (7500-3000BP), with a particular focus on the engravings of the Messak Plateau in southwest Libya. Taking an anthropological approach, the art is examined within the context of the lives of its creators - transhumant cattle-herders who occupied the plateau seasonally. Drawing from fieldwork in Libya together with data from multiple expeditions in the Sahara, the study addresses a major lacuna in Saharan research by focusing on materially constituted, as-lived dimensions at the micro scale. A fundamental but archaeologically elusive aspect of lived experience is a consideration of 'movement', both physical and esoteric. Its incorporation is central to this project, forming a multi-aspected theoretical framework and a methodological tool. Augmented by input from specialists in geomorphology, pastoralism, stone sculpting and animal behaviour, this movement-driven focus has produced a more developed picture of the Messak herder lives, advancing our understanding of these particular non-text, somatic societies. A singular contribution is the creation of a hypothetical model for small-scale, quotidian pastoralist practices, which expands upon the archaeological evidence, fleshing out details of a well-systematised form of dairy pastoralism involving controlled breeding and the processing of milk products. At the same time it is proposed that the herders' relationship with their cattle was one of partnership rather than ownership, involving trans-species empathy and a valuation of animal personhood. This viewpoint is part of a broader set of animal-human relations reflecting a cosmological order that diverges from modern, Western ontological constructs. Other significant findings include detailed information on the role and identity of the image-maker, revisionist data on the amount of effort and skill expended in carving processes, and an examination of the ways in which rock-art was used to manifest social emotional concerns. These were expressed via animal emotions portrayed in the rock-art, and also through performative, gestural markings associated with the imagery. Such expressions include apotropaic, supplicatory or other interactions involving communication with unseen powers.

Description logics (DLs) are knowledge representation formalisms with numerous applications and well-understood model-theoretic semantics and computational properties. SRIQ is a DL that provides the logical underpinning for the semantic web language OWL 2, which is the W3C standard for knowledge representation on the web. A central component of most DL applications is an efficient and scalable reasoner, which provides services such as consistency testing and classification. Despite major advances in DL reasoning algorithms over the last decade, however, ontologies are still encountered in practice that cannot be handled by existing DL reasoners. Consequence-based calculi are a family of reasoning techniques for DLs. Such calculi have proved very effective in practice and enjoy a number of desirable theoretical properties. Up to now, however, they were proposed for either Horn DLs (which do not support disjunctive reasoning), or for DLs without cardinality constraints. In this thesis we present a novel consequence-based algorithm for TBox reasoning in SRIQ - a DL that supports both disjunctions and cardinality constraints. Combining the two features is non-trivial since the intermediate consequences that need to be derived during reasoning cannot be captured using DLs themselves. Furthermore, cardinality constraints require reasoning over equality, which we handle using the framework of ordered paramodulation - a state-of-the-art method for equational theorem proving. We thus obtain a calculus that can handle an expressive DL, while still enjoying all the favourable properties of existing consequence-based algorithms, namely optimal worst-case complexity, one-pass classification, and pay-as-you-go behaviour. To evaluate the practicability of our calculus, we implemented it in Sequoia - a new DL reasoning system. Empirical results show substantial robustness improvements over well-established algorithms and implementations, and performance competitive with closely related work.

A linguagem natural corresponde ao meio mais convencional de comunicação entre as pessoas. O desejo que os seres humanos possuem de se comunicar com as máquinas é evidenciado por pesquisas, que têm sido realizadas desde o final da década de 70, com o objetivo de ter requisições expressas em linguagem natural executadas pelas máquinas. Alguns trabalhos na literatura têm sido propostos com esse fim, entretanto a maioria deles considera requisições expressas apenas em Inglês. Uma maneira de flexibilizar a utilização de várias línguas em sistemas que utilizam linguagem natural é por meio de uma interlíngua, pois essa é uma representação intermediária e processável por máquina das informações contidas em diversas línguas naturais. O trabalho descrito nesta tese propõe que requisições imperativas em linguagem natural sejam convertidas para a interlíngua UNL (Universal Networking Language) e executadas por meio da ativação dos componentes de software apropriados. Para atingir esse objetivo, este trabalho propõe a Arquitetura OntoMap (Ontology-based Semantic Mapping), que utiliza ontologias para realizar o mapeamento semântico entre UNL e componentes de software e para realizar a busca pelos componentes mais apropriados para executar as requisições. A Arquitetura OntoMap conta com (i) um serviço para converter requisições em linguagem natural para UNL; (ii) uma ontologia de alto nível, chamada Ontologia InterComp (Interlíngua-Componentes), que juntamente com regras e inferência, fornece informações semânticas a respeito dos componentes que podem ser utilizados para executar a requisição; (iii) uma Ontologia de Componentes, que relaciona dados das interfaces dos componentes com informações semânticas do domínio de aplicação dos mesmos; e (iv) um Módulo de Busca que utiliza as informações semânticas inferidas e a Ontologia de Componentes para encontrar os componentes apropriados para executar as requisições expressas em linguagem natural. Este trabalho propõe ainda um processo para utilizar a Arquitetura OntoMap em diversos domínios de aplicação e com diferentes conjuntos de componentes. Esse processo foi instanciado considerando componentes desenvolvidos para o domínio de gerenciamento de cursos / Natural Language is the common way of communication between people. The desire of human beings to communicate with machines is evidenced by research, that has been conducted since the late 70?s, triyng to express requests in natural language that can be executed by machines. However, most of the works that have pursued this goal consider requests expressed only in English. A way to facilitate the use of several languages in natural language systems is by using an interlingua. An interlingua is an intermediary representation for natural language information that can be processed by machines. The work described in this thesis proposes to convert imperative natural language requests into the UNL (Universal Networking Language) interlingua and to execute those requests using the apropriate software components. In order to achieve this goal, this work proposes the OntoMap (Ontology-based Semantic Mapping) architecture. It uses ontologies to perform a semantic mapping between UNL and software componente and to search for software components to execute the requests. The OntoMap architecture is composed by (i) a service to convert natural language requests into UNL; (ii) an upper ontology, named InterComp (Interlingua-Components), that uses inference to provide semantic information about components that could be used to execute the requests; (iii) a Components Ontology that relates the component?s interfaces to semantic information about the application domain of the components; and (iv) a search module that uses the infered information and the Components Ontology to reach the components to execute the requests. This work also proposes a process to help the use of the OntoMap architecture in several application domains using different component sets. This process is intanciated considering compoments developed for the course management domain

A cada dia cresce a importância da integração de informações geográficas, em virtude da facilidade de intercambiar dados através da Internet e do alto custo de produção deste tipo de informação. Com o advento da web semântica, o uso de ontologias para descrever informações geográficas está se tornando popular. Para permitir a integração, um dos estágios no qual muitas pesquisas estão focando é o chamado matching das ontologias geográficas. Matching consiste na medida de similaridade entre os elementos de duas ou mais ontologias geográficas. Estes elementos são chamados de conceitos e instâncias. O principal problema enfrentado no matching de ontologias é que estas podem ser descritas por diferentes pessoas (ou grupos), utilizando vocabulários diferentes e perspectivas variadas. No caso de ontologias geográficas os problemas são ainda maiores, em razão das particularidades da informação geográfica (geometria, localização espacial e relacionamentos espaciais), em função da falta de um modelo para descrição de ontologias geográficas amplamente adotado e, também, porque as ontologias são, muitas vezes, descritas em diferentes níveis de granularidade semântica. Estas particularidades das ontologias geográficas torna os matchers convencionais inadequados para o matching de ontologias geográficas. Por outro lado, os matchers existentes para o domínio geográfico são bastante limitados e somente funcionam para ontologias descritas em um modelo específico. Com o objetivo de superar essas limitações, neste trabalho são apresentados algoritmos e expressões (métricas) para medir a similaridade entre duas ontologias geográficas efetivamente, tanto em nível de instâncias quanto em nível de conceitos. Os algoritmos propostos combinam métricas para medir a similaridade considerando os aspectos não geográficos dos conceitos e instâncias com expressões criadas especificamente para tratar as características geográficas. Além disto, este trabalho também propõe um modelo para ontologia geográfica genérico, que pode servir como base para a criação de ontologias geográficas de forma padronizada. Este modelo é compatível com as recomendações do OGC e é a base para os algoritmos. Para validar estes algoritmos foi criada uma arquitetura de software chamada IG-MATCH a qual apresenta também a possibilidade de enriquecer a semântica das ontologias geográficas com relacionamentos topológicos e do tipo generalização/especialização através da análise de suas instâncias. / Integration of geographic information is becoming more important every day, due to the facility to exchange data through the Internet and the high cost to produce them. With the semantic web, the description of geographic information using ontologies is getting popular. To allow the integration, one of the steps in which many researches are focusing is the matching of geographic ontologies. A matching consists on measuring the similarity of the elements, namely either concepts or instances, of two (or more) given ontologies. The main problem with ontology matching is that the ontologies may be described by different communities, using different vocabularies and different perspectives. For geographic ontologies the difficulties may be even worse, for the particularities of the geographic information (geometry, location and spatial relationships) as well as due to the lack of a widely accepted geographic ontology model, and because the ontologies are usually described at different semantic granularities. The specificities of geographic ontologies make conventional matchers not suitable for matching geographic ontologies. On the other hand, the existing geographic ontology matchers are considerably limited in their functionality and deal with ontologies described in a particular perspective. To overcome the current limitations, in this work we present a number of similarity measurement expressions and algorithms to efficiently match two geographic ontologies, at both the concept and instance-level. These algorithms combine expressions used to assess the similarity of the so-called conventional features with expressions tailor made for covering the geographic particularities. Furthermore, this research also proposes a geographic ontology meta-model to serve as a basis for the development of geographic ontologies in order to standardize their description. This model is compliant with the OGC recommendations and is the basis upon which the algorithms are defined. For the evaluation of the algorithms, a software architecture called IG-MATCH was created with an additional feature of making possible to enrich the geographic ontologies with topological relationships and parent-child relationships by the analysis of the instances.

[pt] A principal dificuldade dentro do domínio B2C está em aumentar a utilidade da WWW para o comércio eletrônico através da melhoria das possibilidades oferecidas ao consumidor. Apesar de a WWW permitir ao comprador ter acesso a uma grande quantidade de informação, obter a informação do fornecedor certo que venda o produto desejado a um preço razoável, pode ser uma tarefa muito custosa. Uma das formas de melhorar essa situação é através do uso de agentes inteligentes de busca de informação, isto é, agentes de compra, que auxiliam os compradores a encontrar produtos de seu interesse. Para que isso ocorra esbarra-se em uma dificuldade inerente à própria WWW: a mistura da linguagem natural, imagens e informação de layout de HTML são uma das maiores barreiras para a automatização do comércio eletrônico, pois a semântica da informação é somente compreensível por seres humanos. Desta forma espera- se conseguir agentes de compra mais eficientes quando associados ao uso de ontologias, e lojas virtuais que tenham anotações especiais que sigam uma ontologia. Nessa dissertação fazemos um estudo sobre as principais tecnologias envolvidas no desenvolvimento de ontologias em Ciência da Computação. Fazemos também um estudo de caso sobre a aplicação de ontologias dentro do domínio de B2C, visando assim avaliar o potencial e as dificuldades existentes para o desenvolvimento desse tipo de aplicação. / [en] The main difficulty associated with the B2C domain is increasing the usefulness of WWW for the electronic trade through the improvement of the services provided to the consumer. Even though the WWW allows the buyer to have access to a great amount of information, to obtain the information from the right supplier that sells the desired product by a reasonable price can be a very expensive task. One of the ways of improving the web functionality is through the use of intelligent agents for search of information, that is, the introduction of purchase agents that aid the buyers to find products of their interest. For that to happen we need to overcome an inherent difficulty of the WWW: the mixture of natural language, images and layout information in HTML is one of the greatest barriers for the automation of the electronic trade, because the semantics of the information is only comprehensible for human beings. To solve this problem we hope to produce more efficient purchase agents by associating them to the use of ontologies, and virtual stores that have special annotations that follow ontologies. In the present dissertation we make a study of the main technologies related to ontologies development in computer science. We also develop a case study about the ontologies application to the B2C domain, seeking in this way to evaluate potential and existing difficulties for the development of this type of application.

Chen, Wenhao. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2010. / Includes bibliographical references (p. 63-70). / Abstracts in English and Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Ontologies and Folksonomies --- p.1 / Chapter 1.2 --- Motivation --- p.3 / Chapter 1.2.1 --- Semantics in Folksonomies --- p.3 / Chapter 1.2.2 --- Ontologies with basic level concepts --- p.5 / Chapter 1.2.3 --- Context and Context Effect --- p.6 / Chapter 1.3 --- Contributions --- p.6 / Chapter 1.4 --- Structure of the Thesis --- p.8 / Chapter 2 --- Background Study --- p.10 / Chapter 2.1 --- Semantic Web --- p.10 / Chapter 2.2 --- Ontology --- p.12 / Chapter 2.3 --- Folksonomy --- p.14 / Chapter 2.4 --- Cognitive Psychology --- p.17 / Chapter 2.4.1 --- Category (Concept) --- p.17 / Chapter 2.4.2 --- Basic Level Categories (Concepts) --- p.17 / Chapter 2.4.3 --- Context and Context Effect --- p.20 / Chapter 2.5 --- F1 Evaluation Metric --- p.21 / Chapter 2.6 --- State of the Art --- p.23 / Chapter 2.6.1 --- Ontology Learning --- p.23 / Chapter 2.6.2 --- Semantics in Folksonomy --- p.26 / Chapter 3 --- Ontology Learning from Folksonomies --- p.28 / Chapter 3.1 --- Generating Ontologies with Basic Level Concepts from Folksonomies --- p.29 / Chapter 3.1.1 --- Modeling Instances and Concepts in Folksonomies --- p.29 / Chapter 3.1.2 --- The Metric of Basic Level Categories (Concepts) --- p.30 / Chapter 3.1.3 --- Basic Level Concepts Detection Algorithm --- p.31 / Chapter 3.1.4 --- Ontology Generation Algorithm --- p.34 / Chapter 3.2 --- Evaluation --- p.35 / Chapter 3.2.1 --- Data Set and Experiment Setup --- p.35 / Chapter 3.2.2 --- Quantitative Analysis --- p.36 / Chapter 3.2.3 --- Qualitative Analysis --- p.39 / Chapter 4 --- Context Effect on Ontology Learning from Folksonomies --- p.43 / Chapter 4.1 --- Context-aware Basic Level Concepts Detection --- p.44 / Chapter 4.1.1 --- Modeling Context in Folksonomies --- p.44 / Chapter 4.1.2 --- Context Effect on Category Utility --- p.45 / Chapter 4.1.3 --- Context-aware Basic Level Concepts Detection Algorithm --- p.46 / Chapter 4.2 --- Evaluation --- p.47 / Chapter 4.2.1 --- Data Set and Experiment Setup --- p.47 / Chapter 4.2.2 --- Result Analysis --- p.49 / Chapter 5 --- Potential Applications --- p.54 / Chapter 5.1 --- Categorization of Web Resources --- p.54 / Chapter 5.2 --- Applications of Ontologies --- p.55 / Chapter 6 --- Conclusion and Future Work --- p.57 / Chapter 6.1 --- Conclusion --- p.57 / Chapter 6.2 --- Future Work --- p.59 / Bibliography --- p.63

Les technologies de l'informatique et de la robotique sont en perpétuelle évolution. S'appuyant sur cette évolution technologique, les systèmes d’aide à l’opérateur restent un domaine de recherche d’actualité. Le principal défi des systèmes de la future génération est d'être "intelligents", sensibles au contexte dans un environnement complexe et imprévisible. Cette thèse entre dans ce cadre et traite de la compréhension dynamique du contexte par un robot évoluant dans un tel environnement. En particulier, elle s'intéresse à la question suivante: comment rendre un robot capable de réagir face aux situations de blocage, imprévues dans son plan d’action initial, pour accomplir l’objectif fixé par l’opérateur ? Dans la littérature, ce problème a été soulevé et résolu en partie en programmant dans le système robotique, certaines des fonctions rendant le robot plus autonome. Cependant, l'intégration de ces fonctions dans un même cadre est manquante et plusieurs recherches dans ce sens sont en cours. Dans nos travaux nous proposons un système supportant une approche complète et générique, qui assure à un robot la capacité d’être conscient de la situation de blocage dans laquelle il se trouve et de comprendre et faire face aux situations de blocage rencontrées. Notre approche, nommée Robot Situation AWareness (RSAW) est inspirée de la notion de Situation Awareness (SA) qui a fait ses preuves dans de nombreux domaines notamment dans l’aviation. Nos principales contributions dans RSAW portent sur la conception d’un cadre sémantique intégrant la capacité de compréhension, fondé sur une représentation des connaissances générique, donnant la possibilité d’appliquer des techniques de raisonnement empruntées aux sciences cognitives. L’intégration de RSAW dans un système robotique a également été étudiée, conçue et mise en œuvre dans un système à couches. Ce système d'expérimentation est le robot SAM (Smart Autonomous Majordomo) doté du système AVISO et développé par le CEA-LIST. Les résultats des expérimentations élaborées dans le cadre des travaux menés dans cette thèse sont concluants et prometteurs / Computer technology and robotics are in perpetual evolution. Based on this technological evolution, the operator support systems remain a topical domain of research. The main challenge for the next generation of systems is to be "intelligent", aware of the context in a complex and unpredictable environment. This thesis is into this framework and addresses the dynamic understanding of the context by a robot evolving in such an environment. In particular, the work is interested in the question: How to make a robot able to react to blocked situations unplanned in its initial action plan to achieve the goal set by the operator?In the literature, this issue was raised and resolved in part by programming in robotic system, some of the features making a robot more autonomous. However, the integration of these functions in one framework is missing and more research in this direction is underway. In our work we propose a system supporting a complete and generic approach that ensures a robot the ability to be aware of the blocking situation in which it is found, to understand and deal with deadlock situations encountered. Our approach, called Robot Situation Awareness (RSAW) is inspired by the notion of Situation Awareness (SA), which has been proven in many areas especially in aviation. Our main contributions in RSAW involve the design of a semantic framework integrating the understanding capacity, based on a generic representation of knowledge and giving the possibility to apply reasoning techniques borrowed from cognitive science. Integrating RSAW in a robotic system has also been studied, designed and implemented in a layer system. This experimental system is the robot SAM (Smart Autonomous Majordomo) with the AVISO system developed by CEA-LIST. The conducted experiments allowed testing of the deductive reasoning in resolving a blocked situation and confirmed the need to resort to analogical reasoning. Another wave of experimentation has taken place to prove the effectiveness of our choices. The results of experiments developed as part of the work in this thesis are successful and promising