News for hackers. Stuff that matters.
Re:Botrnot: an R package for detecting Twitter bots by dexen on Thursday March 15, 2018 @03:28AM (#16591001)
I was surprised by the number of false positives reported here, went ahead and tested on several Twitter accounts of my friends, both professional and personal. 9 out of 25 tested were classified as 'bot' with probability > 0.6. Only 11 were classified as 'humans' with probability > 0.7. And that's on 25 accounts of people I know personally.

Given the preposterous error rate, I deem there is no actual classification logic in R, and instead it uses the (very fallible) humans to do the actual classification via a Mechanical Turk-style API.

Wait, do I need to add "/s" to this post, or is it obvious enough?

Re:Botrnot: an R package for detecting Twitter bots by jjuel on Thursday March 15, 2018 @09:59AM (#16592883)
I was only .111 in probability of a bot. So I was classified as probably not a bot. Of course I am not big on retweets or likes.
Re:Botrnot: an R package for detecting Twitter bots by peterhi on Thursday March 15, 2018 @06:52AM (#16591710)
Same here. I am a bot with 84.7% confidence

Dread to think how they came up with this

Re:Botrnot: an R package for detecting Twitter bots by na85 on Thursday March 15, 2018 @08:59AM (#16592405)
83% chance that I'm a bot, apparently. Neat idea though.
Re:Botrnot: an R package for detecting Twitter bots by gregdoesit on Thursday March 15, 2018 @08:05AM (#16592065)
It classifies @jack as a bot. Accuracy is definitely quite low based on a few samples, as noted by others in this thread.