To play off a bit of David’s post on social justice MOOCs, there seems to be a base need for tools for helping identify evil people on the web. That’s not in a dox type of way but more like a way to guide people in determining if accounts have ill intentions.1 That’s probably a messy description but it was brought about by a post I saw on Facebook last night.
Essentially,2 it was a conversation between two people I know- one black male (in Richmond) and one white male (in Baltimore). It was a passionate conversation. The white guy added a screenshot from a Twitter account (included below) into the mix that seemed to confirm all his greatest fears about what was happening in Baltimore.3
It’s a high emotion situation in lots of ways which never helps people critically evaluate items that confirm their fears but I felt like this account had to be fake. Then I wondered if I could prove that? I did a couple of things4 when I looked at this account which are second nature to me now but which could be built into a tool or made into some sort of guide.
Step one– I looked to see when the account was created. This account, @brothertooturnt, was started on 3/27/15. Recent creation and total dedication to a particular topic are suspicious but not conclusive.
Step two– I included two of tweets that felt the most like deliberate provocations below. There was another one (which you can see in the screenshot above) which has since been taken down because a number of people called him out for using this image from a National Geographic episode.
— Jayrome Williams (@BrotherTooTurnt) April 28, 2015
You can then use that URL in a reverse image search engine to see if that picture shows up on other websites. Google does that (directions here) and there are other services like TinyEye. It was pretty obvious that this image wasn’t related to Baltimore. The child was from Memphis and had been in the news recently5 for having pictures like these on Instagram.
I didn’t find a duplicate for the lean/purple drank picture nor for the Twitter profile photo but there are still a decent amount of warning signs that this guy is trying to make a bad situation worse and his efforts should be seen through that lens.
Now it might be interesting to work with the Twitter API and reverse image lookups to create some tool that would let you do something like this simply by entering a username. There is a Python script called Twittego that might also provide some additional ways to evaluate the connections between suspicious accounts. There may be other stuff as well that exists and could be cobbled together. It may be this kind of process shouldn’t be automated.6 I don’t know. The line between skills and simplicity is one I forever struggle with.
1 That would make a good old school rap album name.
2 I am simplifying to get to the point.
3 I’m not seeking to judge him here. It’s really messy stuff when you’re in the midst of fearing for your child’s safety.
4 Enoch might call them “intellectual moves” although it’s interesting to me that they’re blended with the tool/tech application. My intellectual move might just be a default setting of suspicion.
6 If it is automated, it ought to be clearly explained.