Twitter posts about neighborhood characteristics
Model and data analysis on Tweets from the City of Atlanta
Analysis result visualized in a 3D interactive map
Data are collected using streaming API offered by Twitter. Keywords related to the neighborhood characteristics are identified with the help of local periodicals. Tweets with positive and negative attitude are extracted.
Texts are transformed into a set of numeric vectors to train the machine learning models developed by (tokenization, stop-word filtering, stemming, feature representation) CLiPS
Neighborhood-related Tweets are classified into 5 categories: public safety, transportation, Aesthetics, Shopping & Entertainment and Walkability
Derived neighborhood perception is compared with data avaliable in Atlanta's Neighborhood Quality of Life and Health Portal, which is measured at neighborhood planning unit (NPU) level
Master of Computer Science
Master of Computational Science and Engineering
Master of City and Regional Planning
For nearly 20 years, the Center for Spatial Planning Analytics and Visualization (previously the Center for Geographic Information Systems - CGIS) has been at the forefront of collaborative interdisciplinary research for local, state, national, nonprofit, and private entities accoss the nation and the globe.