Many different voting rules have been proposed in the literature and they can select very different alternatives. This naturally raises the question of whether this diversity in outcomes often occurs. Previous works have shown that the probability that voting rules agree on the same outcome is generally quite low under impartial culture. In this article, we use a similar probabilistic approach on single-peaked cultures, which are more structured and typically more realistic than impartial culture. We provide conditions for voting rules to agree under standard single-peaked cultures, and show that the probability of agreement between rather large families of voting rules is much higher under such cultures, with fast convergence of this probability with respect to the number of voters. We finally provide some insights on other structured preference distributions, observing that many exhibit similar convergence in agreement, including the Mallows’ distribution. Our study reveals a tendency of several well-known voting cultures to bias the outcome of voting rules, which is worth knowing before conducting experiments on synthetic data.

We consider the problem of poll manipulation in political elections. In the context of strategic voting, we are interested in whether a polling institute can manipulate the information it communicates to voters in order to influence the outcome of the election. We start with a version of the problem where the polling institute is allowed to send any score to voters. Then, for realistic reasons, we investigate a restricted version in which the polling institute cannot announce scores which are too far from the truthful ones. While we show that both decision problems are computationally hard, we go beyond this worst-case complexity analysis by using probabilistic tools to address the possibility of successful and efficient manipulation in practice, w.r.t. several natural preference distributions.

In order to foster inclusive education, automatic systems that can adapt textbooks to make them accessible to children with Developmental Coordination Disorder (DCD) are necessary. In this context, we propose a task to classify exercises according to their DCD adaptation type. We introduce a challenging exercise dataset extracted from French textbooks, with two major difficulties: limited and unbalanced, noisy data. To set a baseline on the dataset, we use state-of-the-art models combined through early and late fusion techniques to take advantage of text and vision/layout modalities. Our approach achieves an overall accuracy of 0.802. However, the experiments show the difficulty of the task, especially for minority classes, where the accuracy drops to 0.583.

International audience

We consider the problem of the conjoint selection and allocation of projects to a population of agents, e.g. students are assigned papers and shall present them to their peers. The selection can be constrained either by quotas over subcategories of projects, or by the preferences of the agents themselves. We explore fairness and optimality issues and refine the analysis of the rank-maximality and popularity optimality concepts. We show that they are compatible with reasonable fairness requirements related to rank-based envy-freeness and can be adapted to select globally good projects according to the preferences of the agents.

We introduce a way of reasoning about preferences represented as pairwise comparative statements, based on a very simple yet appealing principle: cancelling out common values across statements. We formalize and streamline this procedure with argument schemes. As a result, any conclusion drawn by means of this approach comes along with a justification. It turns out that the statements which can be inferred through this process form a proper preference relation. More precisely, it corresponds to a necessary preference relation under the assumption of additive utilities. We show the inference task can be performed in polynomial time in this setting, but that finding a minimal length explanation is NP-complete.

We consider decision situations in which a set of points of view (voters, criteria) are to sort a set of candidates to ordered categories (Good/Bad). Candidates are judged good, when approved by a sufficient set of points of view; this corresponds to NonCompensatory Sorting. To be accountable, such approval sorting should provide guarantees about the decision process and decisions concerning specific candidates. We formalize accountability using a feasibility problem expressed as a boolean satisfiability formulation. We illustrate different forms of accountability when a committee decides with approval sorting and study the information that should be disclosed by the committee.

We address the problem of multicriteria ordinalsorting through the lens of accountability, i.e. theability of a human decision-maker to own a recommendationmade by the system. We put forward anumber of model features that would favor the capabilityto support the recommendation with a convincingexplanation. To account for that, we designa recommender system implementing and formalizingsuch features. This system outputs explanationsdefined under the form of specific argumentschemes tailored to represent the specific rules ofthe model. At the end, we discuss possible andpromising argumentative perspectives.

Multicriteria analysis is a set of mathematical tools and methods allowing the comparison of different alternatives according to many criteria, often conflicting, to guide the decision maker towards a judicious choice. Multicriteria methods are used in spatial context to evaluate and compare spatial decision alternatives, often modeled through constraint‐based suitability analysis and represented by point, line, and polygon features or their combination, and evaluated on several space‐related criteria, to select a restricted subset for implementation. Outranking methods, a family of multicriteria methods, may be useful in spatial decision problems, especially when ordinal evaluation criteria are implied. However, it is recognized that these methods, except those devoted to multicriteria classification problems, are subject to computational limitations with respect to the number of alternatives. This paper proposes a framework to facilitate the incorporation and use of outranking methods in geographical information systems (GIS). The framework is composed of two phases. The first phase allows producing a planar subdivision of the study area obtained by combining a set of criteria maps; each represents a particular vision of the decision problem. The result is a set of non‐overlapping spatial units. The second phase allows constructing decision alternatives by combining the spatial units. Point, line and polygon feature‐based decision alternatives are then constructed as an individual, a grouping of linearly adjacent or a grouping of contiguous spatial units. This permits us to reduce considerably the number of alternatives, enabling the use of outranking methods. The framework is illustrated through the development of a prototype and through a step‐by‐step application to a corridor identification problem. This paper includes also a discussion of some conceptual and technical issues related to the framework.