This is the social media tool I am building to aid emergency responders and planners. So far, I have used Python and the library Tweepy to do manual searches of strings, filtered by location (usually). This is a short summary of what HazardSearch is intended to do:
- Mining social media sites for information pertaining to extreme events
- Dynamic search algorithm
- Machine learning
- Redistribution and sharing
- To create ‘standardized’ data to use in real time or to piece situations back together… to create better response plans
Here is a diagram that roughly illustrates the process. It was created using Xmind, which was recommended to me by my friend Gray Hayes. This diagram will be modified many times while I further research and improve my understanding of the situation.
As of 4/13/2012:
Original chart on dry erase board
A similar project (and inspiration) is Ushahidi (ushahidi.com)
This page and its subpages are under constant revision.