From spatial data warehouse and decision-making tool to SOLAP generalisation approach for efficient road risk analysis

Plague of modern times, the phenomenon of the traffic accidents hampers heavily resources of the society. The road mows every year thousands of deaths and tens of thousands wounded. This frightening number of victims requires a study of this phenomenon in order to extract all indicators affecting this problem and allow proposing preventive measures. Traditionally, the analysis is performed on alphanumeric data, but some phenomenon as road risk are closely related to their geographic environments where the need to integrate the geographical dimension in the analysis process. In previous work, we have integrated geographic information (GI) represented by a road object in the analysis of this phenomenon. A map with road network theme is so displayed as a response to a SOLAP query. Graphic markers are used to emphasise certain information like dangerous roads. Although displayed maps may be congested and imperceptible, we so propose to improve the displayed information quality by adapting the displayed map to the needs of decision-makers. Maps with different levels of detail are so provided and associated with different analysis levels. This is achieved through on-the-fly map generalisation, which is a well known concept, widely addressed in cartographic domain that we adapt to SOLAP applications. A geospatial decision-making solution based on a spatial data warehouse and a geospatial tool is developed. Some functionalities of our developed geo-decisional tool are described in the experiment section.