One of my favorite things about my job is coming across interesting articles in all the science mags (now that I think about it, it's mostly Nature...). This latest one was a sociological study about online hate groups on social media across the globe...and serious societally-relevant content aside, it was adorable~. ^^
The authors used a unit called a 'cluster' to quantify and organize the hate groups they monitored. and likened the behaviors of the clusters to biological and chemical processes.
For example, when large clusters suddenly die (through banning or what-have-you), the remnants of those clusters and the smaller clusters undergo 'nuclear fusion' into a new successor cluster. It's just a lot of little cute analogies like that~. If you have any experience reading papers, I'd encourage you to give it a look; I'm probably not explaining it all that well.
ANYWAY, the cutest part was the suggestions they gave at the end. They haven't actually investigated/tested any of these; their paper was more about understanding the issue. But here's how they think one might go about managing it:
Policy 1: Ban smaller clusters, not larger ones
Rationale: You knock out more individual members without encouraging the formation of successor clusters, and prevent future large clusters from forming.
I guess it makes some sense, but I can see this looking bad PR-wise...
Policy 2: Ban cluster members randomly, a few at a time
Rationale: You prevent accusations of persecution/free-speech suppression and you don't need any user-specific information, so there's less privacy invasion.
Urrgh I don't like this one either...it feels a bit toothless, not to mention morally questionable...
Policy 3: Organize anti-hate clusters
Rationale: They will fight off and suppress the hate clusters, like a 'human immune system'.
This one just made me laugh. ^^; How on earth do they expect that to work...? I mean, what does the average 'good' user do when they meet a KKK member online, for instance? In my experience, they either ignore/block and move on, or they start bickering with them, both of which accomplish nothing for the health of the social media landscape.
Policy 4: Pit hate clusters of differing ideologies against each other
Rationale: The hate groups will battle among themselves, weakening each other and lessening their overall impact on everyone else.
Oooh, gettin' sneaky~. Even though this is the most outlandish and difficult-to-implement strategy they propose, of all 4, I think it would probably be the most effective. XD It's like with sound waves: if you add one with high positive amplitudes to one with high negative amplitudes, you end up with a wave that's basically zero amplitude, because they cancel each other out.
So what do you think? Do you like any of the researchers' suggestions? Can you come up with better ones?