Organizations try to gain competitive advantages, and to increase customer satisfaction. To ensure the quality and efficiency of their business processes, they perform business process management. An important part of process management that happens on the daily operational level is process controlling. A prerequisite of controlling is process monitoring, i.e., keeping track of the performed activities in running process instances. Only by process monitoring can business analysts detect delays and react to deviations from the expected or guaranteed performance of a process instance. To enable monitoring, process events need to be collected from the process environment. When a business process is orchestrated by a process execution engine, monitoring is available for all orchestrated process activities. Many business processes, however, do not lend themselves to automatic orchestration, e.g., because of required freedom of action. This situation is often encountered in hospitals, where most business processes are manually enacted. Hence, in practice it is often inefficient or infeasible to document and monitor every process activity. Additionally, manual process execution and documentation is prone to errors, e.g., documentation of activities can be forgotten. Thus, organizations face the challenge of process events that occur, but are not observed by the monitoring environment. These unobserved process events can serve as basis for operational process decisions, even without exact knowledge of when they happened or when they will happen. An exemplary decision is whether to invest more resources to manage timely completion of a case, anticipating that the process end event will occur too late. This thesis offers means to reason about unobserved process events in a probabilistic way. We address decisive questions of process managers (e.g., "when will the case be finished?", or "when did we perform the activity that we forgot to document?") in this thesis. As main contribution, we introduce an advanced probabilistic model to business process management that is based on a stochastic variant of Petri nets. We present a holistic approach to use the model effectively along the business process lifecycle. Therefore, we provide techniques to discover such models from historical observations, to predict the termination time of processes, and to ensure quality by missing data management. We propose mechanisms to optimize configuration for monitoring and prediction, i.e., to offer guidance in selecting important activities to monitor. An implementation is provided as a proof of concept. For evaluation, we compare the accuracy of the approach with that of state-of-the-art approaches using real process data of a hospital. Additionally, we show its more general applicability in other domains by applying the approach on process data from logistics and finance. / Unternehmen versuchen Wettbewerbsvorteile zu gewinnen und die Kundenzufriedenheit zu erhöhen. Um die Qualität und die Effizienz ihrer Prozesse zu gewährleisten, wenden Unternehmen Geschäftsprozessmanagement an. Hierbei spielt die Prozesskontrolle im täglichen Betrieb eine wichtige Rolle. Prozesskontrolle wird durch Prozessmonitoring ermöglicht, d.h. durch die Überwachung des Prozessfortschritts laufender Prozessinstanzen. So können Verzögerungen entdeckt und es kann entsprechend reagiert werden, um Prozesse wie erwartet und termingerecht beenden zu können. Um Prozessmonitoring zu ermöglichen, müssen prozessrelevante Ereignisse aus der Prozessumgebung gesammelt und ausgewertet werden. Sofern eine Prozessausführungsengine die Orchestrierung von Geschäftsprozessen übernimmt, kann jede Prozessaktivität überwacht werden. Aber viele Geschäftsprozesse eignen sich nicht für automatisierte Orchestrierung, da sie z.B. besonders viel Handlungsfreiheit erfordern. Dies ist in Krankenhäusern der Fall, in denen Geschäftsprozesse oft manuell durchgeführt werden. Daher ist es meist umständlich oder unmöglich, jeden Prozessfortschritt zu erfassen. Zudem ist händische Prozessausführung und -dokumentation fehleranfällig, so wird z.B. manchmal vergessen zu dokumentieren. Eine Herausforderung für Unternehmen ist, dass manche Prozessereignisse nicht im Prozessmonitoring erfasst werden. Solch unbeobachtete Prozessereignisse können jedoch als Entscheidungsgrundlage dienen, selbst wenn kein exaktes Wissen über den Zeitpunkt ihres Auftretens vorliegt. Zum Beispiel ist bei der Prozesskontrolle zu entscheiden, ob zusätzliche Ressourcen eingesetzt werden sollen, wenn eine Verspätung angenommen wird. Diese Arbeit stellt einen probabilistischen Ansatz für den Umgang mit unbeobachteten Prozessereignissen vor. Dabei werden entscheidende Fragen von Prozessmanagern beantwortet (z.B. "Wann werden wir den Fall beenden?", oder "Wann wurde die Aktivität ausgeführt, die nicht dokumentiert wurde?"). Der Hauptbeitrag der Arbeit ist die Einführung eines erweiterten probabilistischen Modells ins Geschäftsprozessmanagement, das auf stochastischen Petri Netzen basiert. Dabei wird ein ganzheitlicher Ansatz zur Unterstützung der einzelnen Phasen des Geschäftsprozesslebenszyklus verfolgt. Es werden Techniken zum Lernen des probabilistischen Modells, zum Vorhersagen des Zeitpunkts des Prozessendes, zum Qualitätsmanagement von Dokumentationen durch Erkennung fehlender Einträge, und zur Optimierung von Monitoringkonfigurationen bereitgestellt. Letztere dient zur Auswahl von relevanten Stellen im Prozess, die beobachtet werden sollten. Diese Techniken wurden in einer quelloffenen prototypischen Anwendung implementiert. Zur Evaluierung wird der Ansatz mit existierenden Alternativen an echten Prozessdaten eines Krankenhauses gemessen. Die generelle Anwendbarkeit in weiteren Domänen wird examplarisch an Prozessdaten aus der Logistik und dem Finanzwesen gezeigt.

Les humains communiquent via différents types de canaux: les mots, la voix, les gestes du corps, des émotions, etc. Pour cette raison, un ordinateur doit percevoir ces divers canaux de communication pour pouvoir interagir intelligemment avec les humains, par exemple en faisant usage de microphones et de webcams. Dans cette thèse, nous nous intéressons à déterminer les émotions humaines à partir d’images ou de vidéo de visages afin d’ensuite utiliser ces informations dans différents domaines d’applications. Ce mémoire débute par une brève introduction à l'apprentissage machine en s’attardant aux modèles et algorithmes que nous avons utilisés tels que les perceptrons multicouches, réseaux de neurones à convolution et autoencodeurs. Elle présente ensuite les résultats de l'application de ces modèles sur plusieurs ensembles de données d'expressions et émotions faciales. Nous nous concentrons sur l'étude des différents types d’autoencodeurs (autoencodeur débruitant, autoencodeur contractant, etc) afin de révéler certaines de leurs limitations, comme la possibilité d'obtenir de la coadaptation entre les filtres ou encore d’obtenir une courbe spectrale trop lisse, et étudions de nouvelles idées pour répondre à ces problèmes. Nous proposons également une nouvelle approche pour surmonter une limite des autoencodeurs traditionnellement entrainés de façon purement non-supervisée, c'est-à-dire sans utiliser aucune connaissance de la tâche que nous voulons finalement résoudre (comme la prévision des étiquettes de classe) en développant un nouveau critère d'apprentissage semi-supervisé qui exploite un faible nombre de données étiquetées en combinaison avec une grande quantité de données non-étiquetées afin d'apprendre une représentation adaptée à la tâche de classification, et d'obtenir une meilleure performance de classification. Finalement, nous décrivons le fonctionnement général de notre système de détection d'émotions et proposons de nouvelles idées pouvant mener à de futurs travaux. / Humans communicate via different types of channels: words, voice, body gesture, emotions …etc. For this reason, implementing these channels in computers is inevitable to make them interact intelligently with humans. Using a webcam and a microphone, computers should figure out what we want to tell from our voice, gesture and face emotions. In this thesis we are interested in figuring human emotions from their images or video in order to use that later in different applications. The thesis starts by giving an introduction to machine learning and some of the models and algorithms we used like multilayer perceptron, convolutional neural networks, autoencoders and finally report the results of applying these models on several facial emotion expression datasets. We moreover concentrate on studying different kinds of autoencoders (Denoising Autoencoder , Contractive Autoencoder, …etc.) and identify some limitations like the possibility of obtaining filters co-adaptation and undesirably smooth spectral curve and we investigate new ideas to address these problems. We also overcome the limitations of training autoencoders in a purely unsupervised manner, i.e. without using any knowledge of task we ultimately want to solve (such as predicting class labels) and develop a new semi-supervised training criterion which exploits the knowledge of the few labeled data to train the autoencoder together with a large amount of unlabeled data in order to learn a representation better suited for the classification task, and obtain better classification performance. Finally, we describe the general pipeline for our emotion detection system and suggest new ideas for future work.

The use of distributed systems, involving multiple components, has become a common industry practice. However, modeling the behaviour of such systems is a challenge, especially when the behavior consists of several collaborations of different parties, each involving possibly several starting (input) and ending (output) events of the involved components. Furthermore, the global behavior should be described as a composition of several sub-behaviours, in the following called collaborations, and each collaboration may be further decomposed into several sub-collaborations. We assume that the performance of the elementary sub-collaborations is known, and that the performance of the global behavior should be determined from the performance of the contained elementary collaborations and the form of the composition. A collaboration, in this thesis, is characterized by a partial order of input and output events, and the performance of the collaboration is defined by the minimum delays required for a given output event with respect to an input event. This is a generalization of the semantics of UML Activities, where all input events are assumed to occur at the same time, and all output events occur at the same time. We give a semantic definition of the dynamic behavior of composed collaborations using the composition operators for control flow from UML Activity diagrams, in terms of partial order relationships among the involved input and output events. Based on these semantics, we provide formulas for calculating the performance of composed collaborations in terms of the performance of the sub-collaborations, where each delay is characterized by (a) a fixed value, (b) a range of values, and (c) a distribution (in the case of stochastic behaviours). We also propose approximations for the case of stochastic behavior with Normal distributions, and discuss the expected errors that may be introduced due to ignoring of shared resources or possible dependencies in the case of stochastic behaviours. A tool has been developed for evaluating the performance of complex collaborations, and examples and case studies are discussed to illustrate the applicability of the performance analysis and the visual notation which we introduced for representing the partial-order relationships of the input and output events.

No description available.

Apesar de sua inerente indecidibilidade e do problema da negação, extensões da lógica de primeira ordem tem se mostrado capazes de superar a questão da monotonicidade, vindo a constituir esquemas de representação de conhecimento de expressividade virtualmente universal. Resta entretanto solucionar ou pelo menos amenizar as conseqüências do problema do controle, que limitam o seu emprego a aplicações de pequeno a médio porte. Investigações nesse sentido [BOW 85] [MON 88] indicam que a chave para superar a explosão inferencial passa obrigatoriamente pela estruturação do conhecimento, de modo a permitir o exercício de algum controle sobre as possíveis derivações dele decorrentes. O modelo das hiperredes [GEO 85] parece atingir tal objetivo, dado o seu elevado potencial de estruturação e o instrumental que oferece para o tratamento de construções descritivas, operacionais e organizacionais. Além disso, a simplicidade e uniformidade sintática de suas entidades primitivas possibilita uma interpretação semântica bastante clara do modelo original, por exemplo, baseada em grafos. O presente trabalho representa uma tentativa de associar a programação em lógica ao formalismo das hiperredes, visando obter um novo modelo capaz de preservar as expressividade da primeira, beneficiando-se simultaneamente do potencial heurístico e estrutura do segundo. Inicialmente procura-se obter uma noção clara da natureza do conhecimento e de seus mecanismos com o objetivo de caracterizar o problema da representação de conhecimento. Diferentes esquemas correntemente empregados para esse fim (sistemas de produções, redes semânticas, sistemas de frames, programação em lógica e a linguagem Krypton) são estudados e caracterizados do ponto de vista de sua expressividade, potencial heurístico e conveniência notacional. A programação em lógica é objeto de um estudo em maior profundidade, sob os enfoques modelo-teorético e prova-teorético. Sistemas de programação em lógica - particularmente a linguagem Prolog e extensões em nível meta - são investigados como esquemas de representação de conhecimento, considerando seus aspectos sintáticos e semânticos e a sua retação com Sistemas Gerenciadores de Bases de Dados. O modelo das hiperredes é apresentado introduzindo-se, entre outros, os conceitos de hipernodo, hiperrelação e protótipo, assim como as propriedades particutares de tais entidades. A linguagem Hyper, para o tratamento de hiperredes, é formalmente especificada. Emprega-se a linguagem Prolog como formalismo para a representação de Bases de Conhecimento estruturadas segundo o modelo das hiperredes. Sob tal abordagem uma Base de Conhecimento é vista como um conjunto (possivelmente vazio) de objetos estruturados ou peças de conhecimento, que por sua vez são classificados como hipernodos, hiperrelações ou protótipos. Um mecanismo top-down para a produção de inferências em hiperredes é proposto, introduzindo-se os conceitos de aspecto e visão sobre hiperredes, os quais são tomados como objetos de primeira classe, no sentido de poderem ser valores atribuídos a variáveis. Estuda-se os requisitos que um Sistema Gerenciador de Bases de Conhecimento deve apresentar, do ponto de vista da aplicação, da engenharia de conhecimento e da implementação, para suportar efetivamente os conceitos e abstrações (classificação, generalização, associação e agregação) associadas ao modelo proposto. Com base nas conclusões assim obtidas, um Sistema Gerenciador de Bases de Conhecimento (denominado Rhesus em alusão à sua finalidade experimental é proposto e especificado, objetivando confirmar a viabilidade técnica do desenvolvimento de aplicações baseadas em lógica e hiperredes. / In spite of its inherent undecidability and the negation problem, extensions of first-order logic have been shown to be able to overcome the question of the monotonicity, establishing knowledge representation schemata with virtuatLy universal expressiviness. However, one still has to solve, or at Least to reduce the consequences of the control problem, which constrains the use of Logic-based systems to either small or medium-sized applications. Investigations in this direction [BOW 85] [MON 88] indicate that the key to overcome the inferential explosion resides in the proper knowledge structure representation, in order to have some control over possible derivations. The Hypernets Model [GEO 85] seems to reach such goat, considering its high structural power and the features that it offers to deal with descriptive, operational and organizational knowledge. Besides, the simplicity and syntactical uniformity of its primitive notions allows a very clear definition for its semantics, based, for instance, on graphs. This work is an attempt to associate logic programming with the hypernets formalism, in order to get a new model, preserving the expressiveness of the former and the heuristic and structural power of the latter. First we try to get a clear notion of the nature of knowledge and its main aspects, intending to characterize the knowledge representation problem. Some knowledge representation schemata (production systems, semantic networks, frame systems, Logic programming and the Krypton Language) are studied and characterized from the point of view of their expressiveness, heuristic power and notational convenience. Logic programming is the subject of a deeper study, under the model-theoretic and proof-theoretic approaches. Logic programming systems - in particular the Prolog Language and metateuel extensions- - are investigated as knowledge representation schemata, considering its syntactic and semantic aspects and its relations with Data Base Management Systems. The hypernets model is presented, introducing the concepts of hypernode, hyperrelation and prototype, as well as the particular properties of those entities. The Hyper language, for the handling of h y pernets, is formally specified. Prolog is used as a formalism for the representation of Knowledge Bases which are structured as hypernets. Under this approach a Knowledge Brie is seen rrG a (possibly empty) set of structured objects, which are classified as hypernodes, hyperreLations or prototypes. A mechanism for top-down reasoning on hypernets is proposed, introducing the concepts of aspect and vision, which are taken as first-class objects in the sense that they could be (-Ysigned as values to variables. We study the requirements for the construction of a Knowledge Base Management System from the point of view of the user's need-1', knowledge engineering support and implementation issues, actually supporting the concepts and abstractions (classification, generalization, association and aggregation) rYsociated with the proposed model. Based on the conclusions of this study, a Knowledge Base Management System (called Rhesus, refering to its experimental objectives) is proposed, intending to confirm the technical viability of the development of applications based on logic and hypernets.

L'avenir de la communication est perçu comme étant quasiment exclusivement constitué de nœuds mobiles évoluant dans un réseau mobile. Dans ce contexte, différentes approches contribuent continuellement à l'amélioration directe ou indirecte des délais d'acheminement des informations échangées entre les utilisateurs, dont :- l'analyse des traces;- l'évaluation des performances ;- les services de localisation;- le routage. Nous décrivons chacun de ses thèmes et proposons des solutions faisant évoluer l'état de l'art. Celles-ci prennent appui sur des méthodes et outils tels que :- les Réseaux de Petri, pour l'analyse des traces ;- les modèles de mobilité, pour l'évaluation des performances;- l'introduction du social dans les services de localisation;- la mise en place d'une nouvelle métrique pour le routage. Nous montrons comment ces solutions concourent de façon complémentaire les unes avec les autres, à améliorer l'expérience de l'utilisateur. / Future of communication is perceived as being almost exclusively composed of mobile nodes operating in a mobile network. In this context, different approaches contribute to continually improve directly or indirectly, delivery times of information exchanged between users, including:- Trace analysis,- Performance evaluation,- Location services,- Routing.We describe each of these topics and propose solutions by changing the state of the art.These are supported by tools and methods such as:- Petri Nets for Trace analysis,- Mobility Models for Performance evaluation,- Social addings in Location services,- The establishment of metrics for Routing.We show how these solutions work together in a complementary manner with each other, to improve the user experience.

L'engorgement des services des urgences est une question d'actualité. L'augmentation de la fréquentation de ces services par des cas non urgents surcharge ces centres de soins. Par conséquence, la qualité du service délivré aux patients se dégrade et les coûts des soins augmentent. Afin de résoudre ce problème, cette étude porte un intérêt particulier au réseau amont de prise en charge des urgences (pompiers, ambulances privées,médecin libéral, SAMU…). Il s'agit de savoir comment le dispositif actuel doit s'adapter afin de mieux répondre aux besoins des patients. Pour ce faire, le champ de l'étude s'est réparti en deux parties.Une première analyse s'est focalisée notamment sur la régulation médicale au SAMU/Centre15, pivot du réseau amont de prise en charge des urgences et ciment des relations entres les différents acteurs des urgences hospitalières.Une deuxième partie, à vision se situant à un niveau plus global cette fois, prend en compte la phase amont de la prise en charge des urgences dans sa totalité. Elle analyse la complexité du réseau et se centre sur deux éléments d'amélioration possibles, celui de l'introduction des Maisons Médicales de Garde et celui du renforcement de SOS Médecins. / Emergency Department (ED) overcrowding is a topical issue. The increasing attendance at such service by patients with non urgent problems resulted in surcharging the ED. Consequently, the quality of the delivered service is degraded and the care costs are on the increase. In order to solve this problem, a particular interest is taken in the pre-hospital emergency care network (firemen, private ambulances, General Practitioner, SAMU…). Our main purpose is to determine how the current system must adapt to better respond to patients health care needs. This study is divided into two parts: A first analysis focuses on the processing of emergency calls into the SAMU/Centre 15 as itplays an important role in the management of patients flew through this network and guarantees better relations between different actors of this system.A second one deals with the entire pre-hospital emergency care network. The complexity ofthis net is analysed and some alternatives are studied in order to improve the management ofEmergency Department's patients: the establishment of « Maisons Médicales de Garde » and the strengthening of existing medical centres such as « SOS Médecins ».

Les architectures par composants sont de plus en plus étudiées et utilisées pour le développement efficace des applications en génie logiciel. Elles offrent, d’un côté, une architecture claire aux développeurs, et de l’autre, une séparation des différentes parties fonctionnelles et en particulier dans les applications de visualisation scientifique interactives. La modélisation de ces applications doit permettre la description des comportements de chaque composant et les actions globales du système. De plus, les interactions entre composants s’expriment par des schémas de communication qui peuvent être très complexes avec, par exemple, la possibilité de perdre des messages pour gagner en performance. Cette thèse décrit le modèle ComSA (Component-based approach for Scientific Applications) qui est basé sur une approche par composants dédiée aux applications de visualisation scientifique interactive et dynamique formalisée par les réseaux FIFO colorés stricts (sCFN). Les principales contributions de cette thèse sont dans un premier temps, un ensemble d’outils pour modéliser les différents comportements des composants ainsi que les différentes politiques de communication au sein de l’application. Dans un second temps, la définition de propriétés garantissant un démarrage propre de l’application en analysant et détectant les blocages. Cela permet de garantir la vivacité tout au long de l’exécution de l’application. Finalement l’étude de la reconfiguration dynamique des applications d’analyse visuelle par ajout ou suppression à la volée d’un composant sans arrêter toute l’application. Cette reconfiguration permet de minimiser le nombre de services non disponibles. / Component-based approaches are increasingly studied and used for the effective development of the applications in software engineering. They offer, on the one hand, safe architecture to developers, and on the other one, a separation of the various functional parts and particularly in the interactive scientific visualization applications. Modeling such applications enables the behavior description of each component and the global system’s actions. Moreover, the interactions between components are expressed through a communication schemes sometimes very complex with, for example, the possibility to lose messages to enhance performance. This thesis describes ComSA model (Component-based approach for Scientific Applications) that relies on a component-based approach dedicated to interactive and dynamic scientific visualization applications and its formalization in strict Colored FIFO Nets (sCFN). The main contributions of this thesis are, first, the definition of a set of tools to model the component’s behaviors and the various application communication policies. Second, providing some properties on the application to guarantee it starts properly. It is done by analyzing and detecting deadlocks. This ensures the liveness throughout the application execution. Finally, we present dynamic reconfiguration of visual analytics applications by adding or removing on the fly of a component without stopping the whole application. This reconfiguration minimizes the number of unavailable services.

Esta tese apresenta mecanismos computacionais concebidos para reconhecer e inferir estados de ânimo de alunos em interação num ambiente virtual de aprendizagem (AVA). A pesquisa, tendo por base investigações interdisciplinar – Educação, Psicologia Cognitiva e Computação Afetiva (CA) –, apresenta, a título de problema central, o de como alcançar tais objetivos. A Psicogenética Construtivista, base teórica que fundamenta o exame dos aspectos afetivos envolvidos nos processos de aprendizagem, responde pela concepção epistemológica interacionista do AVA ROODA, ambiente de aplicação da pesquisa. A dimensão afetiva é examinada sob a perspectiva cognitivista associada à teoria de appraisal, em especial no que diz respeito ao modelo de Scherer, que tem por finalidade distinguir os diferentes fenômenos afetivos. No que tange à CA, área em torno da qual vêm sendo desenvolvidos estudos que buscam inferir, avaliar e tomar decisões sobre o estado cognitivo-afetivo do aluno, dois aspectos são considerados. De um lado, a análise de como se processa a interação do aluno com o sistema em função de sinais obtidos do comportamento observável no ambiente. De outro, como proceder ao reconhecimento da subjetividade a partir de materiais textuais. Projetos experimentais de CA vêm evidenciando o quanto, no âmbito cognitivo, o reconhecimento da afetividade assume novos contornos. A hipótese de estudo, a de que os estados de ânimo de um aluno podem ser reconhecidos em AVA através de um modelo computacional que inter-relacione fatores de personalidade e de comportamento, pôde ser confirmada, para isso concorrendo o emprego da técnica de redes bayesianas. A opção pelo fenômeno afetivo estado de ânimo, embasada na revisão da literatura, justifica-se por constituir um dos aspectos afetivos mais representativos no ambiente escolar, com respeito ao qual cumpre considerar duas classes de estados de ânimo: animação e satisfação. A Roda dos Estados Afetivos é o espaço de representação para o mapeamento de ambas as classes. Tendo em conta a interdisciplinaridade da pesquisa, e no intuito de identificar os estados de ânimo nas interações de alunos em AVA, construiu-se uma nova funcionalidade para o AVA ROODA, denominada ROODAafeto. Identificar os estados de ânimo em AVA significa ampliar o leque de comunicação professor-aluno, de modo a que não somente os aspectos cognitivos, mas também os afetivos sejam considerados. Em síntese, através dessa funcionalidade, pretende-se proporcionar meios, a partir dos quais seja possível obter informações associadas à vivência afetiva do aluno, notadamente em relação ao ensino não presencial. / This dissertation presents computing mechanisms conceived in order to get to know and infer students’ mood states when interacting in a virtual learning environment (VLE). The research, based on interdisciplinary studies – Education, Cognitive Psychology, and Affective Computing (AC) –, presents, as its main research problem, the aim of reaching such objectives. Constructivist psychogenetics, theoretical foundation for the examination of affective aspects involved in learning processes, is responsible for the interactionist epistemological conception of the ROODA VLE, the application environment for the research. The affective dimension is examined under the cognitive perspective associated to the appraisal theory, mainly regarding Scherer’s model, which aims to distinguish the different affective phenomena. Concerning AC, area that comprises studies developed in order to infer, assess, and take decisions about the student’s affective-cognitive status, two aspects are considered. On the one hand, the analysis of how the interaction of the student with the system is processed was carried out, based on signals obtained from the observable behavior in the environment. On the other hand, it was analyzed how to proceed when recognizing the subjectivity from textual materials. AC experimental projects have evidenced how much, in the cognitive sphere, the recognition of affectivity assumes new forms. The study hypothesis that a student’s mood state can be recognized in a VLE by means of a computing model that inter-relates personality and behavior factors could be confirmed, and for this purpose a Bayesian network technique was used. The option for the affective phenomena mood state, based on the literature review, is justified by the fact that it is one of the most representative affective aspects in the educational environment. With this regard, two types of mood states are considered: being interested and being satisfied. Wheel of Afffective States is the space of representation for the mapping of both types. Considering the interdisciplinarity of the research, and with the purpose of identifying the mood states in the students' interactions in a VLE, a new functionality for the ROODA VLE, called ROODAafeto was developed. Identifying mood states in VLE means to amplify the possibility of teacher-student communication so that affective aspects are also considered besides cognitive aspects. In short, the use of this functionality aims to provide means to obtain information associated with the student’s affective experience, mainly in relation to non-presential education. / Esta tesis presenta mecanismos computacionales diseñados para reconocer e inferir los estados de ánimo de los estudiantes en la interacción en un entorno virtual de aprendizaje (AVA). La investigación, basada en estudio interdisciplinar - Educación, Psicología Cognitiva y Computación Afectiva (CA) - presenta, como problema central, de qué modo alcanzar estos objetivos. La Psicogenética Constructivista, base teórica que se basa en el examen de los aspectos afectivos implicados en los procesos de aprendizaje, responde por la concepción epistemológica interaccionista de AVA ROODA, ambiente de ejecución de la investigación. La dimensión afectiva es examinada desde la perspectiva cognitivista asociada con la teoría de appraisal, en particular en lo que respecta al modelo de Scherer, que tiene por finalidad distinguir los diferentes fenómenos afectivos. Desde el punto de vista de la CA, la zona en torno a los cuales se han desarrollado estudios que buscan inferir, evaluar y tomar decisiones acerca del estado cognitivo-afectivo del alumno, se consideran dos aspectos. Por un lado, el análisis de cómo se procesa la interacción del alumno con el sistema en función de las señales obtenidas del comportamiento observado en el ambiente. Por otra parte, como proceder al reconocimiento de la subjetividad desde los materiales textuales. Proyectos experimentales en CA han confirmando cuanto, en la magnitud cognitiva, el reconocimiento de la afectividad asume nuevos contornos. La hipótesis del estudio, en la que los estados de ánimo de un alumno puede ser reconocida en AVA mediante un modelo computacional que interrelacione factores de la personalidad y del comportamiento, se pudo confirmar, para tanto concurriendo el empleo de la técnica de redes bayesianas. La elección por el fenómeno afectivo estado de ánimo, embasada en la revisión de la literatura, se justifica por constituir uno de los aspectos afectivos más representativos en el entorno escolar, por lo que respecta considerar dos clases de estados de ánimo: animación y satisfacción. El Círculo de los Estados Afectivos es el espacio de representación para el mapeo de ambas clases. Teniendo en cuenta la interdisciplinariedad de la investigación, y el objetivo de identificar los estados de ánimo en las interacciones de los alumnos en AVA, se construyo una nueva funcionalidad para AVA ROODA, llamada ROODAafeto. Identificar los estados de ánimo en AVA significa una ampliación del rango de comunicación profesor-alumno, de modo que no sólo los aspectos cognitivos, sino también los afectivos se tienen en cuenta. En resumen, por medio de esta funcionalidad, se intenta proporcionar los medios, a partir de los cuales sea posible obtener información relacionada con la experiencia afectiva de los alumnos, especialmente en relación con la enseñanza no presencial.

Lagoas costeiras são ecossistemas cruciais para conservação da biodiversidade local e global. O sistema do rio Tramandaí (SRT) apresenta uma série de lagoas costeiras rasas conectadas por canais naturais, e uma saída, via estuário de Tramandaí, para o oceano. Esse sistema é dividido em dois subsistemas, ao norte e ao sul do estuário, resultado de padrões hidrodinâmicos e da influência do vento e que definem características ambientais distintas a esses subsistemas. Um gradiente decrescente, de norte a sul, no estado trófico dessas lagoas já foi verificado. No presente estudo é testada hipótese de variabilidade espacial (lagoas e subsistemas) e temporal (estações do ano) na magnitude de variáveis abióticas e de descritores ecológicos das assembleias de peixes (composição, riqueza, diversidade, dominância) do SRT. Também é avaliada a relação entre variáveis abióticas e descritores das assembleias nesses subsistemas. É comparada a seletividade de captura de um amostrador ativo (rede de arrasto de praia) e de um amostrador passivo (redes de emalhe). Amostragens foram conduzidas de novembro de 2008 a abril de 2010 com rede de arrasto de praia e redes de emalhe, na zona litoral das lagoas Itapeva, Quadros e Malvas no subsistema norte, e nas lagoas Fortaleza, Rondinha e Barros no subsistema sul. Entre abril de 2009 e março de 2010, amostras de água foram coletadas e dez variáveis foram mensuradas em laboratório. Análises de ordenação (NMDS) mostraram clara diferença de composição, abundância numérica (CPUEn) e em biomassa (CPUEb) de espécies entre amostradores. A análises de ANOSIM e SIMPER mostraram baixa similaridade desses parâmetros entre amostradores. Houve diferença no tamanho dos peixes coletados com redes de arrasto e emalhe. Turbidez e DBO5 tiveram médias superiores no subsistema norte, enquanto que condutividade, salinidade, dureza, sólidos totais dissolvidos e transparência de Secchi no subsistema sul. Temporalmente foram verificadas diferenças de temperatura, pH, salinidade, turbidez, dureza, sólidos totais dissolvidos, oxigênio dissolvido, sólidos suspensos e profundidade, mas somente os dados de pH, condutividade, turbidez e DBO5 mostraram interação entre os fatores espaço e tempo. As assembleias das margens rasas da zona litoral foram dominadas por dez espécies, resultando em alta similaridade de composição, CPUEn e CPUEb entre os subsistemas. A riqueza de espécies variou espacial e temporalmente. Diversidade e equitabilidade espacialmente, e CPUEn e CPUEb temporalmente. As assembleias das regiões pelágicas da zona litoral foram dominadas por sete espécies, com dominância numérica das espécies L. anus e A. aff. fasciatus no subsistema norte. A NMDS mostrou similaridades maiores de composição, CPUEn e CPUEb entre amostras de um mesmo subsistema, no entanto, a ANOSIM não apresentou resultados significativos para rejeição das hipóteses nulas. CPUE n e b, diversidade e equitabilidade variaram espacialmente, enquanto a riqueza, temporalmente. A análise de correlação canônica (CCA) mostrou que temperatura, DBO5, sólidos suspensos e dureza responderam por 46% da variabilidade dos dados de abundância numérica das espécies. Os resultados sugerem que amostradores ativos e passivos reproduzem diferentes imagens da assembleia por eles amostrada, indicando que o delineamento amostral para estudos com peixes nesses ecossistemas deve contemplar tais diferenças. As diferenças ambientais dos subsistemas refletem em diferenças na qualidade da água e os resultados obtidos confirmam os pressupostos do gradiente trófico. A estrutura das assembleias de peixes foi mais variável em resposta as condições ambientais em escala espacial, enquanto que temporalmente, houve maior coerência na variabilidade dos dados. As relações das espécies com as variáveis ambientais são discutidas com base no comportamento alimentar e/ou reprodutivo das mesmas e na qualidade do habitat. Tais resultados podem ser úteis como subsídio a elaboração de estudos de manejo costeiro do SRT, sobretudo com relação aos recursos pesqueiros, e na proposição de modelos de estrutura das assembleias de peixes de lagoas costeiras subtropicais. / Coastal lakes ecosystems are crucial for conserving biodiversity both locally and globally. The river Tramandaí system (SRT) comprises a series of shallow coastal lagoons, connected by natural channels, and has an output via Tramandaí estuary into the ocean. This system is divided into two subsystems, north and south of the Tramandaí estuary, as result of hydrodynamic patterns, influence of wind and environmental characteristics that define the subsystems. A trophic gradient decreasing from north to south in those lagoons was identified. In this study is tested the hypothesis of spatial (lagoons and subsystems) and temporal (seasons) variabilities in the amplitude of abiotic variables and descriptors of the fish assemblages (species composition, richness, diversity, dominance) in the SRT. It also assessed the relationship between abiotic variables and descriptors of the fish assemblages in these subsystems. Is compared the selectivity of the active fishing (beach seining) and passive (gill netting) fishing. Sampling was conducted from November 2008 to April 2010 with beach seine and gill nets in the littoral zone of lakes Itapeva, Quadros and Malvas in north subsystem, and Fortaleza, Rondinha and Barros lakes, in south subsystem. Between April 2009 and March 2010 water samples were collected and ten variables were measured in the laboratory. Analyses of ordination (NMDS) showed clear differences in composition, numerical abundance (CPUEn) and biomass (CPUEb) of species between samplers. The analysis of similarity (ANOSIM and SIMPER) showed low similarity of these parameters between nets. There was a difference in the size of the fish caught with beach seine and gill nets. Turbidity and BOD5 averages were higher in the northern subsystem, while the conductivity, salinity, hardness, total dissolved solids and Secchi transparency in the south subsystem. Temporally were observed differences in temperature, pH, salinity, turbidity, hardness, total dissolved solids, dissolved oxygen, suspended solids and depth, but only the data for pH, conductivity, turbidity, BOD5 showed interaction between the factors space and time. The assemblages of the shallow margins of the coastal zone were dominated by ten species resulting in high similarity of composition, CPUEn and CPUEb between subsystems. Species richness varied spatially and temporally, while diversity and evenness only spatially and CPUEn and CPUEb only temporally. The assemblages of the pelagic regions of the coastal zone were dominated by seven species with numerical dominance of L. anus and A. aff. fasciatus in north subsystem. The NMDS showed greatest similarity in composition, CPUEn and CPUEb between samples of the same subsystem, the ANOSIM however, showed no significant results for rejection of the null hypothesis. CPUE n and b, diversity and equitability varied between lakes and subsystems, while the richness between seasons. The canonical correlation analysis (CCA) showed that four variables (temperature, BOD, suspended solids and hardness) account for 46% of data variability of species abundance. Samples of the north subsystem were directly influenced by temperature, BOD and suspended solids while the south subsystem by the hardness. The results suggest that active and passive samplers reproduce different images of the assemblage sampled and indicate that the sampling design for studies of these ecosystems should include such differences. The environmental differences in subsystems reflect in water quality differences and the results confirm the assumptions of a trophic gradient decreasing from north to south in the SRT. The structure of the fish assemblages was more variable in response to environmental conditions in spatial scale between subsystems, while there was greater consistency in the temporal variability of the data. The relationships between the species and environmental variables are discussed based on the feeding behavior and / or reproduction of the same and by habitat quality. These results may be useful as an aid to preparing studies of coastal management of SRT, especially with respect to fisheries, and in proposing models of the structure of the fish assemblages of subtropical coastal lagoons.