To finish off the semester long traineeship at Digital Society School I created a board game that emulates the function of the automated agent our programming team created to test intervention strategies to combat abuse and harassment on Twitter.
The board game includes the strategies we used as well as the obstacles we encountered during development. For example, our initial test at using a bot to interact with Twitter users was met with a suspension of our developer account.
The most interesting part of the game, according to participants at the showcase, is identifying abuse from a pile of ambiguous examples. We found that during development we didn’t always agree when a tweet was truly abusive. It’s an interesting and entertaining part of our game, and a very important part of our research.
The game also includes the counterspeech language we use in our automated interventions. One conversation is neutral, and the other is empathetic.
