Linking GIScience and spatial epidemiology at the 10th Summer School for Spatial epidemiology, social media, and urban health

HeiGIT/GIScience member Sven Lautenbach acted both as part of the boards of directors and as an instructor at the 10th Summer School for Spatial epidemiology, social media, and urban health that took place at the Humboldt University in Bielefeld. This year’s summer school aimed at the use of social media to capture spatial patterns in mental health related outcomes.

The theoretic part of the course covered the topics of urban health, epidemiological challenges and opportunities to urban mental health and the use of OSM data in this context. In addition, Markus Reichert (KIT) presented the outcomes of the psychogeography project the GIScience group is involved as well.

The practical lectures used twitter data recorded before, during and after Hurricane Sandy hit New York city together with R to identify spatial clusters of emotions and to identify driving factors of the number of tweets related to different emotions expressed in the tweets. SatScan, geographically weighted generalized linear models and spatial eigenvector mapping were used for that purpose.

SatScan Clusters for the peri disaster 2 weeks time slot for the number of tweets associated with the emotion disgust.
SatScan Clusters for the peri disaster 2 weeks time slot for the number of tweets associated with the emotion disgust.
Geti's G* for the peri disaster 2 weeks time slot for the number of tweets associated with the emotion disgust.
Geti's G* for the emotion disgust during the two week peri disaster period.

Posted

in

,

by

Tags: