The anonymity that the internet provides makes some people do or say stuff they would not say in the real world. This has led cyberbullying to happen all the time on all social media. How can social media prevent such content or behavior?
The thing about social media is that you can say anything but social media does not have to show your content to people who follow you. An immediate fix that comes to my mind is to “mute” cyberbullying content. We need to detect if a sentence (or image) is bullying someone or a group of people. Our methodology can make mistakes too, so I suggest an exponentially punishing algorithm.
To detect these kinds of social media posts, we can leverage services like Amazon Mechanical Turk to label a huge database of social media posts. We can use deep learning to train a neural network that detects cyberbullying materials. If a post is flagged and the user has published such content for the first time, we would warn (and still mute the post) the user. The user can report the decision as faulty, and then we would check it by hand.
I think content that can reach a huge number of people should be correct (fact-checked) and not fall under cyberbullying. Cyberbullying can make people physically hurt themselves, which is a terrible thing to happen. Content can be offensive to people (to some extent). Offensive content is different from cyberbullying content.
Vulnerable individuals are usually isolated from other people, and social media is their only source of social content. This is why I think social media should prevent cyberbullying, and it is a good thing.
Muting cyberbullying content will prevent the sharing of such content. Bad content might still be published, but given the fact that we have reduced them, it will result in a narrower audience, which might prevent a vulnerable person from seeing the content and hurting themselves.