IGU Presentation: Spatio-Temporal Text Mining
Your browser does not support iframes. Try Firefox.
I recently presented some ongoing work on trying to identify anchor points (areas that are regularly visited by people such as workplace, home, the pub, etc.) at the IGU 2013 conference in Leeds. The most interesting part of the work, so far, is using a GIS algorithm to try to identify areas that have a large enough number of messages posted in them to be considered an anchor point.