Identifying Protective Decision Making Behavior through GeoLocated Tweets

The goal of this project is to identify potential evacuation behavior based on geotags and timestamps within a Twitterer's contextual stream (EPIC).

This project is a continuation of work previously done with Andrew Hardin and Ellie Falletta in CU Geography 5303 (Spring 2014)

---

Top Pages

• Latest Statistics
• Known Weaknesses

Full Contents

• Jul 10, 2015 Data Cleaning and Complications
• Jul 7, 2015 Movement Path Analysis for 97 Users
• Jul 6, 2015 What contextual looks like
• Jun 17, 2015 Summary of Data Filtering To-Date
• Apr 14, 2015 Mapping Qualitative Data with Timeline
• Feb 11, 2015 Geographical Distribution of Tweets & Users
• Dec 16, 2014 New Hierarchical Coding Scheme
• Dec 9, 2014 December 09 Progress Meeting
• Dec 1, 2014 Mapping Qualitative Data
• Oct 17, 2014 New Cluster Evacuation Analysis
• Sep 13, 2014 What questions can we ask of the data?
• Sep 10, 2014 Qualitative Coding
• Sep 8, 2014 September Status
• Sep 1, 2014 Known Weaknesses
• Aug 27, 2014 Summary Statistics
• Aug 25, 2014 Data Statistics
• Aug 25, 2014 New Data Structure
• Aug 22, 2014 A New Algorithm
• Aug 20, 2014 Triangles (Previous Analysis Method)
• Aug 20, 2014 Previous (triangles) Design
• Aug 20, 2014 Final Workflow
• Aug 19, 2014 First Pass at Qualitative Analysis
• Aug 15, 2014 Next Steps
• Aug 6, 2014 Newest Statistics
• Aug 4, 2014 New Filters Applied to Twitterers
• Aug 1, 2014 Proper Clustering Algorithm for Tweets
• Jul 28, 2014 Triangle Analysis

---

References

1. Project EPIC at University of Colorado Boulder
2. Sandy-GIS Project from GIS 3 at University of Colorado Boulder (Spring 2014)