Twitter's Gender Representativeness Evaluation

UCGIS Poster Presentation for evaluating the gender representativeness of twitter

View Full Image

There have been emerging interests in location-based social media in recent decades, and twitter is one of the most popular platforms integrated with cross-disciplinary research. However, the representational bias of twitter data is under investigation, and a good understanding of such issue remains an ignorable gap. Focusing on the proportions of major genders, this research explores the representativeness of twitter data in Clarke County, Georgia, by comparing the demographics of active Twitter users with such characteristics of the local population at census tract level, further generating the maps of gender representativeness of Twitter. The results of spatial and non-spatial characteristics are also visualized using bivariate choropleth map with corresponding plot chart.

Click to view the poster gallery at UCGIS Symposium 2020.