News
Release 2 of the dataset is available.
The submission deadline has been extended to January 31, 2007.

Pandemic diseases such as avian influenza cause extremely infectious disease outbreaks. Pandemic influenza viruses have demonstrated their ability to spread worldwide within months or even weeks, and to cause infections in all age groups. While the ultimate number of infections, illnesses, and deaths is unpredictable, and could vary tremendously depending on multiple factors, it is nonetheless certain that without adequate planning and preparation, an influenza pandemic has the potential to overwhelm current public health and medical care capacities at all levels. Controlling the spread of a pandemic requires early detection via appropriate surveillance, along with implementation of appropriate responses (e.g., isolation of cases, quarantine of contacts, antiviral drug treatment and prophylaxis).

Meeting these demands for pandemic detection and response poses numerous significant challenges for data mining. We must develop and model a time-varying spatial-social network capturing collocation and effective contact patterns. We must conduct model-based data aggregation to identify the onset of disease and other qualitative indicators of disease spread. We must identify critical individuals and critical locations, in order to support targeted vaccination and targeted detection goals (respectively).

The goal of the SDM-07 Workshop on Data Mining for Pandemic Preparedness is to stimulate research within and across these areas. The workshop aims to bring together researchers from relevant sub-disciplines of data mining (addressing spatial, temporal, and social aspects of the problem), simulation and stochastic modeling, and active sampling and control. Workshop participants are encouraged to center their contributions on a dataset from a comprehensive pandemic simulation. The goal is not to have a "bake-off", but rather to establish a point of contact unifying many different research angles in data mining. Researchers working on the data mining challenges posed by pandemic preparedness are encouraged to submit papers and contribute to the workshop discussion.

Synthetic dataset covers 100 days, with 1.6 million people at 240,000 locations based on the city of Portland, Oregon. (left) 3D view of Portland with one location and its data highlighted. (right) 2D map with points for home (blue) and work (red) locations.