Spatio sequential patterns

S2P is a tool developed in collaboration with IRSTEA and the University of New Caledonia in the context of health data mining. It is a web application available for health professionals to help them to understand and visualize large amounts of epidemiological data. Indeed, the management of health risks such as epidemics causes the production of large amounts of spatio-temporal data. For example, health professionals have at their disposal information by geographic area (for a neighborhood in a city) such as the number of dengue cases, the weather over a  time... These characteristics, both spatial and temporal, require the development of new methods to fully exploit the richness of the data.

The application is based on a map divided into geographical areas and a spatial sequential patterns viewer associated with a geographical area. The map allows you to visualise the organisation of the regions. Each zone is identified by a name and a contour. The user may at any time move, zoom in on different areas and choose one of them.

The second visualisation enable to list the patterns associated with a zone. A spatio-temporal pattern is represented by a graph. The red node correspond to dengue cases. The dotted edges correspond to temporal informations (« after ») and the dotted edges correspond to geographical informations  (« near »).

Spatio sequential patterns

S2P has been published in PAKDD 2012.

Last update on 20/01/2015