The international Network Against Cyber Hate (INACH) combats discrimination on the Internet. INACH unites organizations around the world, incorporating different cultural backgrounds and speaking many languages. It has a diverse perspective, allowing the network to counter and address all forms of online discrimination. INACH adds value to the Internet and brings the online in line with Human Rights.
INACH trains various organizations, leaders and individuals around the world on detecting and responding to cyber hate. Critical to this effort is the development of tools for training that keep up with the use of the internet and forms of hate occurring in various environments. INACH approach Samurai to develop a tool that pinpoints live violence on the internet as an integral aspect of their training program.
Samurai partnered with INACH to create cutting edge tools monitoring aggressive and violent communities on Reddit and on the dark web in places including 4chan. Samurai’s objective was to provide experts in training with live examples of various types of hate speech.
CHALLENGE AND RESULT
Samurai created the ability to distinguish violence from harsh language that is commonplace in places like 4chan. Removing false positives from real violence is difficult for AI, and is why incorporating 4chan into training previously had been challenging. Samurai further used its ability to distinguish types of violence with fine grained precision to support the broad efforts of the education programs.
By working with Samurai, INACH was able to incorporate case work into their training involving listening to carefully filtered conversations to surface speech that is most concerning and aligned with training objectives. Participants in the training were served with actual examples of violence to address with counter-speech, rather than sifting through a high percentage of comments with strong but non-violent language which was not the priority for INACH and distracting and time consuming for program participants.