PhD Student –
- Subject: Geovisualization for decision support in a context of uncertainty: application to mountain victim location
- Keywords: Geovisualization, Decision support, Uncertainty, Spatial reasoning
- Funding: ANR CHOUCAS
- Start: October 1st, 2018
Mountain victim search is based on the analysis of a set of clues provided by the requester (person who triggers rescue), describing geographical objects (landscape elements, landmarks, followed routes…). This analysis is based on a visual analysis of cartographic supports of different kinds, possibly supplemented by other documentary media or external information from community sites (geoweb data…).
This is part of a spatial reasoning process, which consists of formulating hypotheses based on the first information provided by the requester and spatial landmarks, refining them progressively through dialogue and analysis of spatial indices, and then deducing the most likely location of the victim. The process of locating victims is essentially based on the expert knowledge of the rescuers and a hypothetico-deductive approach that can be strongly tainted with uncertainty.
Today, few tools are available to support rescuers in their geovisual analysis and help them explore different location scenarios. We hypothesize that a multidimensional geovisualization environment can support spatial reasoning and be a decision-making aid, since it is capable of implementing various visualization methods (cartographic and non-cartographic), which integrate the representation of the uncertainty and can be convened according to the context of the research, the assumptions made and the deductions emitted by the user.