L’étude de la conflictualité armée dans les pays en développement et des effets des interventions extérieures visant à la circonscrire constitue un enjeu majeur et multidisciplinaire. La complexité des processus sociaux entrelacés dans de tels contextes génère des défis particuliers pour les approches de modélisation et simulation informatique. En effet la forme des conflits actuels, centrés autour des populations locales, a amené les forces de stabilisation à développer des actions dites d'influence, qui se donnent pour but de convaincre les locaux de la légitimité des opérations. Ce travail en retient trois (opérations psychologiques OMI/PSYOPS, opérations civilo-militaires et Key Leader Engagements) et propose une modélisation permettant de représenter la propagation de leurs effets au sein d'une population virtuelle réaliste. Cela nécessite une représentation détaillée des mécanismes psychologiques de persuasion et d’influence et des structures sociales considérées, et constitue donc un apport à l’analyse des processus de communication au sein de réseaux sociaux. On fait le choix d'une simulation sociale orientée agents permettant de représenter d'une part un échantillon composé uniquement d'individus (jusqu’à environ 10 000) dotés de mécanismes de traitement réalistes, et d'autre part la complexité des sphères de sociabilité collective en générant cette population sous forme d'un réseau multicouche. Le modèle conceptuel ainsi conçu permet d’intégrer pour chaque étape des critères pertinents pour la simulation des actions, mais aussi, eut égard à la nature même de celles-ci et de leurs enjeux, d’éclairer certains mécanismes sociaux caractéristiques d’un conflit armé / The study, analysis and understanding of armed violence phenomenons in developing countries and of the effects of international military interventions aimed at resolving them is a crucial contemporary issue. These situations coincidentally present, considering the plethora and complexity of the implicated and intertwined social processes, a challenge for social computing, modeling and simulation. A challenge but also an opportunity, as the evolution of the forms of conflict, today centered on the local populations, has prompted the armies tasked with implementing stabilization missions to develop influence actions. Such operations, overall concerned with persuading the locals of the legitimacy of the operations, allow a more comprehensive approach to conflict resolution, beyond the simple use of force. Modeling some of these specific actions (PSYOPS, CIMIC, and Key Leader Engagement) is a credible project and a contribution to analysis of communication and persuasion processes in social networks by taking into account detailed and specific social and cultural factors. This research thus presents a conceptual model allowing simulation of the effects of these specific actions of influence in a realistic civilian population. We chose an agent-based approach as these lend particularly well with this type of research, allowing us to generate a group of up to 10,000 agents, composed solely of individuals for a detailed cognitive treatment, and structured as a multilayer network for representing complex sociality. Given the nature of such actions and their context of application, such a model also highlights some social mechanisms typical of armed conflict situations

Text in English / Interactive health communication between the health-care professional (HCP) and patient relationship for diabetes health-care positively contributes to patient-centred care. Hence individual patient concerns are addressed and catered for in the medical system. The purpose of this study was to analyse in-depth how HCP-patient relationships and HCP-HCP teamwork dynamics positively contribute to effective diabetes patient care and treatment adherence. Different health communication models and theories were reviewed and a conceptual framework was developed from the literature. A qualitative case study approach was used to collect data at Stanger Hospital’s diabetes clinic. Data was collected using three different methods, namely in-depth interviews with HCPs and patients individually, observations conducted at the clinic analysing both the HCPs and patients, and finally, documentation that emerged as a third data collection method where patient files and diabetes educational material were analysed at the clinic. Triangulation by means of the three methods ensured that reliable, valid and credible data was collected in the field. Diabetes health-care and treatment management are affected by the social context/social system which includes family and culture. These social factors are acknowledged as core in the literature. However, a single comprehensive health communication model did not exist solely in this regard. The data indicated that at the Stanger Hospital’s diabetes clinic, patient-centred (individual tailor-made treatment plans) care was only implemented after patients had defaulted treatment for reasons linked to their social circumstances. The findings of the study indicate that teamwork was favoured in the HCP-HCP relationship at the diabetes clinic, and that this made a positive contribution to effective diabetes patient care. HCPs were overburdened at the clinic since patient numbers were high and there were staff shortages. The patients’ empirical data indicated that interactive communication positively contributed to their medical concerns being catered for at the diabetes clinic, but this tended to occur only after non-adherence. Patients required care and support from HCPs in order to learn to accept diabetes and manage their illness. / Sociology / D. Litt. et Phil. (Sociology)