Using Twitter Data to Predict Violent Events
MetadataShow full item record
With the popularity of social media, platforms such as Twitter have become the perfect place for extremist groups to espouse ideas and recruit members. The goal of this project was to determine if the frequency of use of twitter by extremist groups or individuals associated with these groups and ideologies reliably correlates to the lead up to a violent event, such as the shooting perpetrated by Dylann Roof in Charleston. After this shooting, evidence of the perpetrators activity online surfaced. He had a blog called ‘Last Rhodesian’ where he openly talked about his desire to remove black people form his home state. The week directly preceding his attack, he posted a last message about his frustration of other online extremist’s unwillingness to act, and how he had decided to take it upon himself and hoped to inspire others to do the same. After this came to light, the effect of social media on radicalization within the US became a more popular topic, and it was this that ultimately inspired this project. Being able to use readily available information such as twitter data, or other social media platforms, to predict a violent event would give law enforcement a greater chance of preventing the event. Not only would lives potentially be saved if a violent event were prevented, but extremist ideology driven events have the to potential to inspire more of the same, or to inspire others to adopt the same beliefs. Preventing these events from occurring could then prevent the spread of motivation to commit such acts. This project would then hope to determine if a pattern can be found, and whether that pattern could be used to predict future events.