Student Takes On Online Harassment at Penn State Hackathon

While the Internet connects us, opens our eyes to new cultures, and gives us the opportunity for learning far beyond what was possible only a few decades ago, it is hardly a perfect creation. Online harassment is a serious, pervasive, and pernicious problem. From death threats to doxing, the Internet can be a terrifying place — full of hate. 
 But Hack Harassment believes it doesn’t have to be that way. We know the Internet can be better and kinder. So we asked students around the nation to help us find a tech solution to this growing epidemic.
 We are excited to share this story from the winner of the Hack Harassment challenge at
Hack PSU, Amanda Goonetilleke. We hope her words help you realize that it’s possible to end online harassment. Join the movement today!

— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —

As much as I love the internet for making just about anything more accessible and convenient, there is no denying that it is has also become a harmful medium for online harassment. Many people may be targeted online to varying degrees, and some may dismiss it because the internet seems less real. But online harassment is a very real problem and I believe everyone would be hard-pressed to find someone in their life it hasn’t affected.

As a Rutgers student, the saddest illustration of how very real and tragic online harassment can be is the suicide of a fellow Rutgers student, Tyler Clementi, who died September 22, 2010 after being cyber­bullied by his roommate. His death created a national dialogue on cyberbullying and LGBT issues, and I believe there is so much more we can all do to combat these problems.

That’s why, when I first heard about the Hack Harassment challenge at Hack PSU, I immediately fell in love with its initiative to combat online harassment with technology. Although I was initially overwhelmed with the challenge, after talking to Hack Harassment representatives, I became inspired by their cause and began brainstorming how to tackle the problem at hand.

My project was to create a text categorization algorithm to classify whether ​an online interaction’s ​text is harassment or not. I used IBM’s AlchemyAPI to help with the text classification. Their API can detect keywords and their respective relevance and sentiment, as well as provide scores for different emotions in the text. With more concrete data extracted from the text, I could identify whether it was negative or positive, if it targeted a certain group or individual, and depending​ on the related emotions, I could classify whether the text was not harassment, probably harassment, or indeed harassment.

To my surprise, I won the Hack Harassment challenge! I worked on this project by myself in the span of 24 hours, but for other people who also find this initiative as meaningful and interesting as me, I believe there are so many other ways to help hack harassment. Creating tools to block, report, or even prevent harassment would have an incredible impact that could help make the internet a more productive and positive place. Just by attempting this challenge, I learned so much about online harassment. For anyone curious about hacking harassment, my advice is to go for it, and that any technology or tool you develop will be an important and meaningful contribution.

To learn more about how you can contribute, visit and follow us on Twitter and Facebook.