Researchers Scour Twitter for Cyberbullying

Education Week recently reported on a computer program that was designed to search Twitter for signs of cyberbullying. Professor Amy Bellmore and graduate students Junming Sui and Kwang-Sung Jun at the University of Wisconsin at Madison developed the program, which uses “machine learning” to read each tweet and analyze the word usage for signs of bullying. Read their full findings here.

The program was able to find 15,000 bullying-related tweets out of the roughly 250 million tweets that occur during the course of a given day. Researchers aim to use the data in order to recognize signs that point to children who may be in need of adult intervention.

“Kids are pretty savvy about keeping bullying outside of adult supervision, and bullying victims are very reluctant to tell adults about it happening to them for a host of reasons,” said Amy Bellmore. “They don’t want to look like a tattletale, or they think an adult might not do anything about it.”

But Twitter was just a start. The researchers involved in the project conceded that Facebook and other methods of communication not bound by the confines of 140-character tweets hold the capacity for harsher, more extensive cyberbullying.

According to this infographic, one in 10 kids on Facebook report being bullied on the site, which adds up to about 800,000 kids. And let’s also not forget about text messages too. The average U.S. teen sends 60 texts per day, leaving plenty of statistical room open for abusive text messages.

Read Caity’s insight about the problems of cyberbullying in today’s schools.