El trabajo realizado en esta tesis pertenece al área de argumentación en inteligencia artificial. La representación de conocimiento en un formalismo basado en argumentación se realiza a través de la especificación de argumentos, cada uno en favor de una conclusión a partir de ciertas premisas. Dado que estas conclusiones pueden estar en contradicción, se producen ataques entre los argumentos. Luego, la evaluación de toda la información presente podría dar preponderancia a algunos argumentos por sobre aquellos que los contradicen, produciendo un conjunto de conclusiones que se considera ran garantizadas. El objetivo principal de esta tesis es la definición de un nuevo marco argumentativo capaz de manejar dinámica de conocimiento. En este sentido, se da una representación no sólo a los argumentos, sino que también se introduce la noción de evidencia como entidades especiales dentro del sistema. En cada instante, el conjunto de evidencia se corresponde con la situación actual, dándole contexto al marco argumentativo. La plausibilidad de los argumentos en un instante dado depende exclusivamente de la evidencia disponible. Cuando la evidencia es suficiente para dar soporte a un argumento, éste se denominará activo. También se considera la posibilidad de que algunos argumentos se encuentren activos aun sin encontrar soporte directamente desde la evidencia, ya que podrían hacerlo a través de las conclusiones de otros argumentos activos. Estas conexiones entre argumentos dan lugar a lo que en esta tesis se denomina estructura argumental, proveyendo una visión un tanto más compleja que la usual en cuanto a la representación de conocimiento argumentativo. Los resultados obtenidos en esta tesis permitirán estudiar la dinámica de conocimiento en sistemas argumentativos. En la actualidad, ya se han publicado artáculos que presentan un formalismo que combina argumentación y la teoría clásica de revisión de creencias. En esta línea de investigación se denen operadores de cambio que se aplican sobre el marco argumentativo abstracto dinámico y tienen como objetivo alcanzar cierto estado del sistema; por ejemplo, garantizar un argumento determinado. Por otra parte, este marco también permitirá estudiar métodos para acelerar el computo de garantía a partir del proceso de razonamiento realizado en estados anteriores.

With the fast development of radio technologies, wireless systems have become more convoluted. This complexity, accompanied by an increase of the number of connections, is translated into a need for more parameters to analyse and decisions to take at each instant. AI comes into play by automating these processes, particularly with Deep Learning techniques, that often show the best accuracy. However, the high performance by these methods also comes with the drawback of behaving like a black box from the view of a human. To this end, eXplainable AI serves as a technique to better understand the decision process of these algorithms. This thesis proposes an eXplainable AI framework to be used on Reinforcement Learning agents, particularly within the use case of antenna resource adaptation for network energy reduction. The framework puts a special emphasis on model adaptation/reduction, therefore focusing on feature importance techniques. The proposed framework presents a pre-model block using Concrete Autoencoders for feature reduction and a post-model block using self-supervised learning to estimate feature importance. Both of these can be used alone or in combination with DeepSHAP, in order to mitigate some of this popular method’s drawbacks. The explanations provided by the pipeline prove useful in order to reduce model complexity without loss of accuracy and to understand the usage of the input features by the AI model. / Med den snabba utvecklingen av radioteknologier har trådlösa system blivit alltmer invecklade. Denna komplexitet, kombinerat med en ökning av antalet anslutningar, innebär att fler parametrar behöver analyseras, och fler beslut behöver fattas vid varje ögonblick. AI kommer in i bilden genom att automatisera dessa processer, särskilt med Deep Learning-tekniker, som ofta uppvisar bäst noggrannhet. Men den höga prestandan med dessa metoder kommer också med nackdelen att tekniken beter sig som en svart låda från en människas synvinkel. Förklarlig AI fungerar därför som en teknik för att bättre förstå beslutet som fattas av dessa algoritmer. Denna avhandling föreslår ett förklarligt AI-ramverk som ska användas inom förstärkningsinlärning, särskilt inom användningsfallet med antenn-resursanpassning för energireduktion i trådlösa nätverk. Det föreslagna ramverket sätter en särskild tonvikt på modellanpassning/modellreduktion. Ramverket innehåller ett förmodellblock som använder Concrete Autoencoders för Feature Reduction och ett post-modellblock som använder självövervakad inlärning för att uppskatta Feature Importance. Båda dessa kan användas ensamt eller i kombination med DeepSHAP, för att lindra några av denna populära metods nackdelar. Feature Importance-uppskattningarna från ramverket visar sig vara användbara för att minska modellkomplexitet utan förlust av noggrannhet och för att förstå användningen av Input Features av AI-modellen.

<p>The thesis describes steps taken towards using graphic representation and visual modelling support for the knowledge acquisition process in knowledge-based systems – a process commonly regarded as difficult. The performance of the systems depends on the quality of the embedded knowledge, which makes the knowledge acquisition phase particularly significant. During the acquisition phase, a main obstacle to proper extraction of information is the absence of effective modelling techniques.</p><p>The contributions of the thesis are: introducing a methodology for user-centred knowledge modelling, enhancing transparency to support the modelling of content and of the reasoning strategy, incorporating conceptualisation to simplify the grasp of the contents and to support assimilation of the domain knowledge, and supplying a visual compositional logic programming language for adding and modifying functionality.</p><p>The user-centred knowledge acquisition model, proposed in this thesis, applies a combination of different approaches to knowledge modelling. The aim is to bridge the gap between the users (i.e., knowledge engineers, domain experts and end users) and the system in transferring knowledge, by supporting the users through graphics and visualisation. Visualisation supports the users by providing several different views of the contents of the system.</p><p>The Unified Modelling Language (UML) is employed as a modelling language. A benefit of utilising UML is that the knowledge base can be modified, and the reasoning strategy and the functionality can be changed directly in the model. To make the knowledge base more comprehensible and expressive, we incorporated visual conceptualisation into UML’s diagrams to describe the contents. Visual conceptualisation of the knowledge can also facilitate assimilation in a hypermedia system through visual libraries.</p><p>Visualisation of functionality is applied to a programming paradigm, namely relational programming, often employed in artificial intelligence systems. This approach employs Venn-Euler diagrams as a graphic interface to a compositional operator based relational programming language. </p><p>The concrete result of the research is the development of a graphic representation and visual modelling approach to support the knowledge acquisition process. This approach has been evaluated for two different knowledge bases, one built for hydropower development and river regulation and the other for diagnosing childhood diseases.</p>

The thesis describes steps taken towards using graphic representation and visual modelling support for the knowledge acquisition process in knowledge-based systems – a process commonly regarded as difficult. The performance of the systems depends on the quality of the embedded knowledge, which makes the knowledge acquisition phase particularly significant. During the acquisition phase, a main obstacle to proper extraction of information is the absence of effective modelling techniques. The contributions of the thesis are: introducing a methodology for user-centred knowledge modelling, enhancing transparency to support the modelling of content and of the reasoning strategy, incorporating conceptualisation to simplify the grasp of the contents and to support assimilation of the domain knowledge, and supplying a visual compositional logic programming language for adding and modifying functionality. The user-centred knowledge acquisition model, proposed in this thesis, applies a combination of different approaches to knowledge modelling. The aim is to bridge the gap between the users (i.e., knowledge engineers, domain experts and end users) and the system in transferring knowledge, by supporting the users through graphics and visualisation. Visualisation supports the users by providing several different views of the contents of the system. The Unified Modelling Language (UML) is employed as a modelling language. A benefit of utilising UML is that the knowledge base can be modified, and the reasoning strategy and the functionality can be changed directly in the model. To make the knowledge base more comprehensible and expressive, we incorporated visual conceptualisation into UML’s diagrams to describe the contents. Visual conceptualisation of the knowledge can also facilitate assimilation in a hypermedia system through visual libraries. Visualisation of functionality is applied to a programming paradigm, namely relational programming, often employed in artificial intelligence systems. This approach employs Venn-Euler diagrams as a graphic interface to a compositional operator based relational programming language. The concrete result of the research is the development of a graphic representation and visual modelling approach to support the knowledge acquisition process. This approach has been evaluated for two different knowledge bases, one built for hydropower development and river regulation and the other for diagnosing childhood diseases.

Les connaissances perceptivo-gestuelles sont difficiles à saisir dans les Systèmes Tutoriels Intelligents. Ces connaissances sont multimodales : elles combinent des connaissances théoriques, ainsi que des connaissances perceptuelles et gestuelles. Leur enregistrement dans les Systèmes Tutoriels Intelligents implique l'utilisation de plusieurs périphériques ou capteurs couvrant les différentes modalités des interactions qui les sous-tendent. Les « traces » de ces interactions –aussi désignées sous le terme "traces d'activité"- constituent la matière première pour la production de services tutoriels couvrant leurs différentes facettes. Les analyses de l'apprentissage ou les services tutoriels privilégiant une facette de ces connaissances au détriment des autres, sont incomplets. Cependant, en raison de la diversité des périphériques, les traces d'activité enregistrées sont hétérogènes et, de ce fait, difficiles à modéliser et à traiter. Mon projet doctoral adresse la problématique de la production de services tutoriels adaptés à ce type de connaissances. Je m'y intéresse tout particulièrement dans le cadre des domaines dits mal-définis. Le cas d'étude de mes recherches est le Système Tutoriel Intelligent TELEOS, un simulateur dédié à la chirurgie orthopédique percutanée. Les propositions formulées se regroupent sous trois volets : (1) la formalisation des séquences d'interactions perceptivo-gestuelles ; (2) l'implémentation d'outils capables de réifier le modèle conceptuel de leur représentation ; (3) la conception et l'implémentation d'outils algorithmiques favorisant l'analyse de ces séquences d'un point de vue didactique. / Perceptual-gestural knowledge is multimodal : they combine theoretical and perceptual and gestural knowledge. It is difficult to capture in Intelligent Tutoring Systems. In fact, its capture in such systems involves the use of multiple devices or sensors covering all the modalities of underlying interactions. The "traces" of these interactions -also referred to as "activity traces"- are the raw material for the production of key tutoring services that consider their multimodal nature. Methods for "learning analytics" and production of "tutoring services" that favor one or another facet over others, are incomplete. However, the use of diverse devices generates heterogeneous activity traces. Those latter are hard to model and treat.My doctoral project addresses the challenge related to the production of tutoring services that are congruent to this type of knowledge. I am specifically interested to this type of knowledge in the context of "ill-defined domains". My research case study is the Intelligent Tutoring System TELEOS, a simulation platform dedicated to percutaneous orthopedic surgery.The contributions of this thesis are threefold : (1) the formalization of perceptual-gestural interactions sequences; (2) the implementation of tools capable of reifying the proposed conceptual model; (3) the conception and implementation of algorithmic tools fostering the analysis of these sequences from a didactic point of view.