My thesis work was related to automatic generation of critical situations in virtual environments. We focused on two particular types of critical situations : dilemmas and ambiguous situations. The challenge of this work is to generate automatically these situations using knowledge models that are not intended to be used for the representation of dilemmas and ambiguous situations.
A dilemma is commonly defined as a situation that includes a difficult choice. It refers to a situation where individuals have to choose between two, or more, inconvenient options. Ambiguity, however, refers to situations that are interpreted in several ways. In our work, we propose a formalization of these two notions, which is inspired by humanities and social sciences. Using this formalization, we propose a set of algorithms and generation techniques which use knowledge models - filled out by domain experts - that are not intended to represent dilemmas and ambiguous situations. The use of these models enables the generation engine to infer new knowledge used to extract the entities (e.g. actions, events, objects) that can potentially produce situations that meet the properties defined in the dilemma and ambiguity formalization. With regard to moral dilemmas generation, in order to propose an adapted content to each learner, it is necessary to take into consideration the values system of each person. Thus, we propose to operationnalize the theory of universal values of Schwartz. With regard to ambiguity generation, it is necessary to take into account the level of knowledge of the learner regarding the world variables. Thus, we propose to model the mental representation of the learner. This representation must take uncertainties into consideration. To do so, we use the belief functions theory as it enables the system to quantify the uncertainty, the conflict and the ignorance.
My thesis was part of the project MacCoy Critical funded by the ANR. This latter aims to study and to improve training systems using simulation and virtual environments in medical education (obstetrics) and in driving education (novice drivers during the first months of autonomous driving). In this project, the approach and the architecture are elaborated from a multidisciplinary viewpoint in order to provide more appropriate and flexible Virtual Environments for Learning/Training to support the acquisition of Non-Technical Skills by Humans.
More information : https://maccoy.hds.utc.fr/