Inspiration
Through direct experience, we are educating the bullies and the people how it feels to be bullied.
What it does
Users will talk to our bot, bully simulator ReginaG to find and learn more effective way to stand up to a bully verbally. Analysis at the end of conversation will determine the level of confidence in user's texts and either suggest a better way or compliment your achievement. After the Analysis, the bot recommends few options where to look for more information on bullying.
How we built it
We have used machine learning to train ReginaG to use certain kind of speech with similar level of average threats. Using Twillio API, we added text-exchangeability to ReginaG. The server runs on AWS. Through numerous tests, we were able to make the bot converse with users. Tone analysis from IBM-Watson cognitive services provided valuable insights on confidence level. We are on Facebook messenger as well!
Challenges we ran into
We have changed our approach few times completely. And once we settled on the current method, one of the challenges we faced was trying to balance the educational point and exposing to full experience of getting bullied.
Accomplishments that we're proud of
It works.
What we learned
We researched different methods of standing up to a bully, how bullies attack victims, and where to look for more help.
What's next for ReginaG
More variety to be added on speech/text.
Log in or sign up for Devpost to join the conversation.