Situation Recognition Using EventShop by Vivek K. Singh & Ramesh Jain

Author:Vivek K. Singh & Ramesh Jain
Language: eng
Format: epub
Publisher: Springer International Publishing, Cham

5.4 Example: Modeling Epidemic Outbreaks

Let us illustrate the process of situation modeling by considering “epidemic outbreaks.” Given as is, “epidemic outbreak” is a vague undefined notion. In fact, not even all experts agree on what constitutes an epidemic. Here, we discuss the workflow for one possible modeling of epidemics. Following Sect. 5.2, we first identify the output state space (i.e., required classification into low, mid, and high risk of outbreak). Next (see Fig. 5.5), we identify the spatiotemporal bounds being considered: the USA, with a spatial resolution of 0.1 latitude X 0.1 longitude and reevaluation to be made every 5 min. The next step is identifying the relevant features which define the situation output. Let us define “epidemic outbreaks” as a classification on “growing unusual activity.” While this is a single feature, it is not atomic (i.e., cannot be derived directly using one data source). Hence, we follow the process recursively and try to model “growing unusual activity.” This feature is defined based on two component features: “unusual activity” and “growth rate.” It turns out that “unusual activity” is also not atomic and needs to be split into the features of “historical activity level” and “growth rate.” Let us assume that the historical activity level is available from a curated database and current activity level can be measured based on the frequency of terms indicating influenza-like illness (ILI) on Twitter stream. Similarly, the growth rate can be measured from the Twitter stream.

Fig. 5.5Base model created for epidemic outbreaks

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.