Laboratoire de Génie Informatique et d’Automatique de l’Artois

Aurélie AKLI

Ph.D. student
Member of the research themes:
Contact information:
    No publication meets the provided criteria.

    Ph.D. topic: "Assistance in reasoning under inconsistencies for uncertainty resolution"


    In information fusion systems [1], managing inconsistent information is a challenging cognitive task, especially when the resulting decision-making stakes are high. For example, in surveillance tasks often supported by command-and-control systems, inconsistencies in the available information can have a crucial impact on situation understanding. In maritime surveillance, these inconsistencies can reflect illicit or malicious activities, distressed vessels, or simply sensor malfunctions. In the field of intelligence, analysts employ certain methodologies to assess hypotheses arising from potentially contradictory observations. In both cases, the operator or analyst’s task is particularly difficult due to the significant amount of information to process, the diversity of sources with varying quality, the often limited time they have to resolve the problem, and the psycho-social pressure they may face.

    The formal framework of belief functions [2] allows for the fusion of information from partially reliable sources, as well as capturing the heterogeneity and imperfection of the information. Moreover, measuring inconsistency has been the subject of several works within this framework (e.g., [3-7]). Each measure of inconsistency satisfies specific mathematical properties that characterize its behavior in line with certain human intuitions. However, characterizing what constitutes inconsistency itself poses a challenge since it heavily depends on the operational context (mission, weather, geopolitics, environment, etc.).

    The major challenge of this work will be to reconcile the measure of inconsistency with human intuition regarding its behavior, while avoiding the reproduction of reasoning biases. The goal will be to identify one or more measures of the degree of inconsistency within the framework of belief functions, with controlled semantics and behavior, adapted to the operational context, and meaningful to the analyst or operator. Knowledge engineering and human-centered system design methodologies will play a central role in eliciting and understanding the reasoning of experts faced with inconsistent information (e.g., [8-10]).

    The developed approach will be tested on real and simulated data from various application domains, which will be specified during the thesis (e.g., maritime surveillance, industrial site monitoring, intelligence, wildfire prevention, etc.).


    ‌[1] D. Dubois, W. Liu, J. Ma, H. Prade. The basic principles of uncertain information fusion. An organised review of merging rules in different representation frameworks. Information Fusion 32:12-39, 2016.
    [2] G. Shafer. A mathematical theory of evidence. Princeton University Press, Princeton, 1976.
    [3] S. Destercke, T. Burger. Toward an axiomatic definition of conflict between belief functions. IEEE Trans. Syst. Man Cybern. B 43(2):585–596, 2013.
    [4] F. Pichon, A.-L. Jousselme, N. Ben Abdallah. Several shades of conflict. Fuzzy Sets and Systems 366(1) : 63-84, 2019.
    [5] F. Pichon, S. Destercke, T. Burger. A consistency-specificity trade-off to select source behavior in information fusion. IEEE Transactions on Cybernetics 45(4):598-609, 2015.
    [6] A.-L. Jousselme, F. Pichon, N. Ben Abdallah, and S. Destercke. A Note About Entropy and Inconsistency in Evidence Theory. Int. Conf. on Belief Functions (BELIEF 2021), vol. 12915 of Lecture Notes in Computer Science, pages 215-223, Springer, 2021.
    [7] N. Ben Abdallah, S. Destercke, A.-L. Jousselme, and F. Pichon. Logical and evidential inconsistencies meet : first steps. Int. Conf. on Belief Functions (BELIEF 2021), vol. 12915 of Lecture Notes in Computer Science, pages 207-214, Springer, 2021.
    [8] A.-L. Jousselme, G. Pallotta, J. Locke. Risk game : Capturing impact of information quality on human belief assessment and decision making. Int. Journal of Serious Games 5(4):23–44, 2018.
    [9] G. A. Boy. Articulating Human Systems Integration. In Design for Flexibility : A Human Systems Integration Approach, pages 29-47, Springer International Publishing, 2021.