S01E05 - The Mosaic Pieces
The picture below show us the cases during the flu pandemic period. What we want to do is find the case at the early stage, at the green arrow. This early detection can lead to early warning and early response so the pandemic it self can be prevented.
A surge of new technologies and social media innovations is altering the media landscape. These chances are affecting the way people behave. Now, the INTERNET has made measurable what was previously immeasurable and as we all know, people do online search for every problem they encounter, so basically everything, including online search before they go to the doctor consulting about a disease.
We find that these online queries are very valuable resources which hold information regarding the possibility of pandemic occurrence. We build our model by filtering the search queries related to flu, and do some advance data mining technique to detect the outbreak. Our method relies on the emerging research on public health field that proven statistically effective describing health activity patterns.
This is our initial algorithm that describe the flu activity during 2009-2010. As we can see, our model fit the real world problem, where the evaluator capture increasing flu activity on May, 2009 (blue line) , and on the middle of July it shown the outbreak happened. As all we know that it is a swine flu stroke the globe. Providing early detection is important, because it could lead to early warning and rapid response from stakeholders.