Also, Just peeked at the last 6 months of false positive autoflags with scores over 156. 25 of the 32 would have been prevented with my score adjustment proposal. :o
@Andy @Undo Can you please restore the deleted answer on stackoverflow.com/c/charcoal/questions/627? It was automatically deleted as a negatively-scored answer when the owner's account was deactivated, but I still feel it's valuable to have for context.
@angussidney Which repo should the issue be filed on? SD, MS, or website?
@Rubiksmoose may I refer you to "if a post is edited by the author in a way which is mostly offensive, then the appropriate feedback is tp- as we would’ve wanted that edit to be blocked by the system." in the feedback guidelines?
A fair interpretation, though as additional context, they also left a similarly profane comment complaining that people were overcomplicating things on one of the answers.
also, they re-edited the profanity back in after I rolled it back :-) (I mod-flagged)
My interpretation may have been fair, but apparently it was also wrong XD
@RyanM I'll take that excuse lol
I'm not sure I've seen anyone say this yet, but @DanielWiddis you are doing some fantastic work with data! I'm going to try to read further into it when I have time, but seriously nice work.
I've been happy to read it too. I hypothesized a few years ago that we were essentially double counting on a few specific reasons. I'd have to find a way to search through thousands of smokey reports in this room to find the conversation, but it was things like Keyword in Body paired with Keyword in Title for the same key word. The data nerd in me is really liking the posts.
The manager nerd is upset there aren't pretty pictures :)
I don't use Mac, but I have coworkers that love postico. No idea on feature set or why it's good, but that's the recommendation I see them give all the new employees
hm, google tells me it's not free so maybe not what you are looking for.
@Undo Tried that. The download from the website won't connect to my DB. The version from the App store lets me browse schema but I can't query without paying for a subscription (not one time fee)
@RyanM ...okay, I lied. I got to 98.9% accuracy on 3952 posts. For context, that's 7.3% of the total FP questions in metasmoke, and 4.7% of FP posts overall.
Also what the heck is BARC? Because I'm guessing they don't mean: "BARC Shelter is not-for-profit, no-kill, privately run animal shelter located in Williamsburg, Brooklyn"
[ SmokeDetector | MS ] Bad keyword in body, blacklisted website in body, pattern-matching product name in body, pattern-matching website in body, potentially bad ip for hostname in body, +2 more (490): This Oil Will Boost Your Immunity by peggya shley on astronomy.SE
@Mast That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body; in addition, the regex contains an unescaped "."; in most cases, it should be "\."; append -force if you really want to do that.
@Mast Indeed. SO is old enough now that this kind of thing happens on occasion: someone makes an account, asks a question or two, leaves, then 10 years later creates a new product and decides to spam it
@Makyen That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body; append -force if you really want to do that.
@Makyen That pattern looks like it's already caught by Blacklisted website in body and Blacklisted website in answer; append -force if you really want to do that.
@DanielWiddis That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body; append -force if you really want to do that.
@DanielWiddis That pattern looks like it's already caught by Potentially bad keyword in answer and Potentially bad keyword in body; append -force if you really want to do that.
