RICHMOND — A map of geotagged tweets containing hate words shows where the most racist and homophobic tweets come from in the United States, a California professor says.
The map was created by Humboldt State University Professor Monica Stephens as part of her research on the geographic origins of online hate speech.
The data behind the map is based off every geocoded tweet in the United States from June 2012 to April 2013 that contains one of the “hate words.” It is built off 150,000 tweets from that time.
According to an online description of the map, HSU students in Stephens’ Advanced Cartography course read each tweet and classified it as positive, neutral or negative based on a predefined rubric.
Only tweets identified by readers as negative are included in the analysis. This allowed the team to exclude key words that were not intended to be negative (e.g. “dykes on bikes #SFPride”).
“Hateful tweets were aggregated to the county level and then normalized by the total number of tweets in each county,” Stephens said. “This then shows a comparison of places with disproportionately high amounts of a particular hate word relative to all tweeting activity.”
For example, Orange County, California has the highest absolute number of tweets mentioning racial slurs. But because Orange County has many, many more tweets than most other counties, hateful tweets are easily lost in the chatter and do not appear prominently on the map.
So the map is really a view of where in the US hateful tweets would be the most prominent.
The map has three categories: racist, homophobic and disability. Under each are subcategories for specific slurs.
When the filters were selected per each category, only one section of Virginia showed up in the red, which indicates the most extreme level of hate. Western parts of the state indicated a high level of racism, based on the map’s analysis.
Click over to see what area of the state are indicated by the map to be most racist.
Read more about the research and methods behind this project visit www.FloatingSheep.org.