The first chapter is based on a paper with Jin Li in fair division. It was recently discovered that on the domain of Leontief preferences, Hurwicz (1972)'s classic impossibility result does not hold; that is, one can find efficient, strategy-proof and individually rational rules to divide resources among agents. Here we consider the problem of dividing l divisible goods among n agents with the generalized Leontief preferences. We propose and characterize the class of generalized egalitarian rules which satisfy efficiency, group strategy-proofness, anonymity, resource monotonicity, population monotonicity, envy-freeness and consistency. On the Leontief domain, our rules generalize the egalitarian-equivalent rules with reference bundles. We also extend our rules to agent-specific and endowment-specific egalitarian rules. The former is a larger class of rules satisfying all the previous properties except anonymity and envy-freeness. The latter is a class of efficient, group strategy-proof, anonymous and individually rational rules when the resources are assumed to be privately owned. The second and third chapters are based on two working papers of mine in decision making under uncertainty. In the second chapter, I study the wealth effect under uncertainty --- how the wealth level impacts a decision maker's degree of uncertainty aversion. I axiomatize a class of preferences displaying decreasing absolute uncertainty aversion, which allows a decision maker to be more willing to take uncertainty-bearing behavior when he becomes wealthier. Three equivalent preference representations are obtained. The first is a variation on the constraint criterion of Hansen and Sargent (2001). The other two respectively generalize Gilboa and Schmeidler (1989)'s maxmin criterion and Maccheroni, Marinacci and Rustichini (2006)'s variational representation. This class, when restricted to preferences exhibiting constant absolute uncertainty aversion, is exactly Maccheroni, Marinacci and Rustichini (2006)'s ariational preferences. Thus, the results further enable us to establish relationships among the representations for several important classes within variational preferences. The three representations provide different decision rules to rationalize the same class of preferences. The three decision rules correspond to three ways which are proposed in the literature to identify a decision maker's perception about uncertainty and his attitude toward uncertainty. However, I give examples to show that these identifications conflict with each other. It means that there is much freedom in eliciting two unobservable and subjective factors, one's perception about and attitude toward uncertainty, from only his choice behavior. This exactly motivates the work in Chapter 3. In the third chapter, I introduce confidence orders in addition to preference orders. Axioms are imposed on both orders to reveal a decision maker's perception about uncertainty and to characterize the following decision rule. A decision maker evaluates an act based on his aspiration and his confidence in this aspiration. Each act corresponds to a trade-off line between the two criteria: The more he aspires, the less his confidence in achieving the aspiration level. The decision maker ranks an act by the optimal combination of aspiration and confidence on its trade-off line according to an aggregating preference of his over the two-criterion plane. The aggregating preference indicates his uncertainty attitude, while his perception about uncertainty is summarized by a generalized second-order belief over the prior space, and this belief is revealed by his confidence order.

Lorsque la vue, qui est un important moyen de communication, est altérée, alors l'acquisition de l'information s'en trouve modifiée, dégradée ou limitée. A l'ère du monde numérique, le Web regorge d'informations réparties sur différents sites et mises en forme par les développeurs et designers. De nombreuses pathologies visuelles peuvent entraîner des difficultés dans l'accès à ces informations. Au-delà même de ces informations, l'accès aux outils et services est lui aussi limité. Des difficultés dans la perception des couleurs, des taches dans le champ visuel ou un champ visuel réduit sont tout autant de sources de difficultés. Chaque personne a une vision qui lui est propre. Chez les personnes qui ont une basse vision, les pathologies donnent des évolutions spécifiques chez chacune d'entre elles. De plus les méthodes de compensation acquises sont différentes d'une personne à l'autre. Des outils d'assistance existent depuis de nombreuses années et tentent de répondre aux besoins des personnes ayant une basse vision en proposant des adaptations visuelles. Les principales limites de ces outils résident notamment dans le fait qu'ils ne sont pas en capacité de prendre en compte les besoins très spécifiques de chaque personne. Ces travaux de recherche se concentrent donc autour de l'analyse des besoins réels des utilisateurs et de l'élaboration d'une nouvelle approche qui se base sur les préférences personnelles de l'utilisateur. L'objectif final est d'automatiser la transformation des pages Web en fonction des préférences propres à un utilisateur pendant qu'il navigue sur le Web. Divers algorithmes ont été utilisés, notamment des algorithmes évolutionnaires, afin de réaliser des compromis entre les préférences de l'utilisateur et l'apparence originale de la page Web. La thèse développe de manière approfondie les principaux problèmes touchant les personnes en situation de basse vision et des éléments sur les modèles de couleurs et de contrastes. Puis elle présente un langage de modélisation des préférences basé sur la logique, une modélisation du problème comme un problème d'optimisation, des algorithmes de résolution, un démonstrateur, et des expérimentations sur des pages Web réelles. / When the sight, which is the main communication way, is altered, then the information acquisition process is also modified, degraded or limited. In today's digital world, the Web is a wealth of information organized by designers and developers and available on different Websites. Many visual pathologies can lead to difficulties in accessing this information. Beyond this information, the access to the different tools and services is also affected. Difficulties in color perception, cloud-like white patches or dark areas in a visual field, or a reduced visual field are all sources of difficulties. Each person has a particular vision. Several persons with the same pathology may even have different visions. Several assistive tools have been proposed that apply visual adaptation, trying to meet the needs of people with low vision. Main limits of these tools are mainly the unability of taking into account the very specific needs of each person. These research works are focused on the real user's needs analysis and on making a new approach based on the personal user's preferences. The final target consists in automatizing the Web page transformation according to the specific preferences of a particular user. This transformation occurs along the navigation from page to page. Different algorithms have been used, especially evolutionary algorithms, in order to make tradeoffs between the user's preferences and the original appearance of the page. The thesis further develops main problems encountered by people with low vision and some notions on color models and contrast relations. After that, we present a preference modeling language based on logics, a modeling of the problem as an optimization problem, some resolution algorithms, a tool and experiments on several real Web pages.