[ SmokeDetector | MS ] URL in title, bad keyword in body, bad keyword in title, blacklisted website in body, blacklisted website in title, repeated URL at end of long post: offersreviews.org/malexpro/ by ticu weba on apple.SE
Well comodo do offer completely functional free products. I'm happily using the free firewall. I haven't noticed any spam on SU before ... but that was the second in a few minutes ... by the same user.
hovering shows "Flagged by avery" and there's a 1 in front of the flag, but I guess this is a known bug from earlier versions, although I had sort of hoped for it to be fixed in beta3
just to be explicit, the flag icon itself has a link attached to it, which is however broken
Chrome (I think it is) obscures this by showing "post/.../flag_logs" with an ellpisis, but if you hover long enough it reveals that the ellipsis expands to "undefined"
the link is there on posts which were not autoflagged, with the same error
Alright smokey, I'm about to type some stuff. Don't interrupt me this time...
Updated Graph, now with even more colors!
This has three trend lines per section. Trend lines show the average time to removal for a reported post.
The solid line is the average trend over that period of time. The dashed line is the time to deletion for posts created between UTC+4 and UTC+12 (spamming hour!). The dotted line is the time to deletion outside of spamming hour.
A couple interesting things that I noticed - time to deletion during spam hour was higher when we didn't cast any automatic flags. It was removed faster outside of spam hour.
That reversed when we started issuing a single auto-flag. The spam hour time to deletion is slightly lower than the average. Comparing the two time periods though, time to deletion during non-spam hour at the end of the non-flagging time period and the end of the single flag period are roughly the same.
Again, the three flags time period is so new that there isn't enough data to draw conclusions yet.
@Andy FYI, as a newcomer to this project, and after reading the description several times, I'm not sure what the first results graph in the meta post is supposed to communicate.
the blue line reflects the number of True Positives, given a specific weight of the number of filters/triggers detected for a given post (and thus the autoflagging thresholds), with the left Y-axis as the scale for it
Right, with you so far. But what's the point? I think the point is that the better the "reasons" are for the auto-flagging, the more accurate the true positive percentage.
But I'm not sure if that's what's intended, or if that's even correct.
How a graph like this might be used as you fine-tune things might also be worth explaining... so you have the power to raise or lower the percentage accuracy of auto-flags by adjusting the minimum "weight" required to trigger them
Nothing on fire... yet. As we all know, the Fires of Chaos and Darkness and Evil pop up in the most random locations and times to cause disruption and damage. :P
like the commit I almost pushed to Smokey that would have torched Smokey completely. :)
so then he would be a smokey SmokeDetector. *da dum tish*
> The above graph plots the weight of the reasons against its overall volume of reports and accuracy. As minimum weight increases, accuracy (yellow line and rightmost Y-axis) and total reports (blue line) on the left-hand scale decrease.
@Glorfindel Yeah, but look at the revision history on that last report. Only 2 hours ago "You're all wrong. I made a free energy device for $10 and open sourced it so people can make their own. Why would you steal my shit and claim it as your own? youtube.com/watch?v=5dZTqiPX0c8 what the fuck is this? What about my micro organism game theory and the results you published. Fuck the lot of you"
