[ SmokeDetector | MS ] Bad keyword in body, blacklisted website in body, body starts with title and ends in url, link at end of body, potentially bad ns for domain in body, +1 more (394): you reversal in aging signs, but such by user1062776 on superuser.com
@SmokeDetector @Makyen moved that expression you watched in the blacklist, since it got 3 TP hits in <30 mins. Wondering, though, if it'd be better to blacklist just keto\W*body. Will it be caught if a post contains ketobodytone?
@double-beep That pattern looks like it's already caught by Bad keyword in body and Bad keyword in answer; append -force if you really want to do that.
[ SmokeDetector | MS ] Url-only title, bad keyword in body, bad keyword in title, body starts with title and ends in url, link at end of body, +7 more (791): health-body.org/keto-one/ by Phan1989 on askubuntu.com
@angussidney + @tripleee I now have a couple of autos. I think the rebase (of weights) calculation that occurs every couple of months (whatever it's called) shifted us (me, I'm certain) under 99.75 causing flagging to be disabled. It wasn't showing on the UI untill I went through it and I wasn't pinged with the "Total accuracy must be >= 99.75%" message. All three of us are AFing normally now. Thanks again for your assistance.
I tried to post something once and I got the error:
Your post could not be submitted at this time due to the volume of spam and abuse originating from your network.
I was very confused because I do not have a business to promote, and I don't recall ever trolling recently. So if this is abou...
@Makyen That pattern looks like it's already caught by Pattern-matching website in answer and Pattern-matching website in body; append -force if you really want to do that.
@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 Pattern-matching website in answer and Pattern-matching website in body; append -force if you really want to do that.
@ThomasWard Thanks. Interesting. I only checked the watchlist, because they didn't qualify for the blacklist, based on their # TP.
The -force was left over from creating the commands prior to updating the watchlist not to detect them in the creative-bio\w+ watch. I should have removed the -force prior to actually running the commands once I changed the order of what I was doing to updating the creative-bio\w+ watch. Ironically, having them watched and blacklisted was what was already happening prior to my changes. Thanks for catching it.
@ThomasWard That might be better, but I was trying to leave it more or less the way I found it, but just eliminate it duplicating what I'd added as specific entries, and the one other specific entry I did find.
Well, it's actually creative-bio\w+\b(?!\.com(?<=creative-biostructure\.com))(?!\.com(?<=creative-biolabs\.com))(?!\.com(?<=creative-biogene\.com))(?!\.com(?<=creative-bioarray\.com)) at this point in order to not duplicate the ones which were added as individual items.
@Makyen Wouldn't it be more prudent to just drop the ones we already catch on and use the wider detection pattern? Because it seems that all the badness ends up being with the wider detection pattern
@ThomasWard While I agree that does qualify for blacklisting, I'm usually a bit leery of blacklisting patterns. I'd generally rather see individual domains blacklisted and leave the watch to pick up anything new, in this case. I just don't feel that creative-bio\w+\.com is really all that specific, given that it's only composed of a common English word followed by "bio". I could certainly see a legit company picking a domain name with that pattern.
@ThomasWard I could be convinced to blacklist, and definitely think it should be at least watched.
Frankly, I like for us to have another blacklist which was blacklisted patterns, rather than keywords.
agree on the broader watch pattern. And then maybe blacklist depending on the number of overall watch triggers. But that's me being overly 'broad' with my analysis of things. I've been busy heh
@thesecretmaster The NS checks do lookups based on the actual domain. The blacklisted website list does not have any bookending added to the regex. IMO, that's bad. Unfortunately, there are patterns in the blacklisted website list that both rely on the existence of no bookending and which expect there to be bookending. We really should go through every single entry and verify which was intended. We could then change that blacklist to include bookending (different than just \b) by default.
@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 Pattern-matching website in body and Pattern-matching website in answer; append -force if you really want to do that.
@Makyen That pattern looks like it's already caught by Pattern-matching website in body and Pattern-matching website in answer; append -force if you really want to do that.
@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.
@thesecretmaster My expectation was that we'd want to run every entry through MS to see the entries where there's a difference in detections between the bare entry as a regex and using a bookended version. We could then concentrate on the entries which have a difference between those two. Those could be updated to catch everything appropriate with bookending applied. After that, we could then have the bookending applied generally.
Yes, obtaining the data can be automated. It was my expectation to have it automated, running each through MS' search capability, with a pause between each, so that it doesn't take up all that much of the processing available to MS.
@Makyen No rush at all, just wanted to offer if that'd speed up the process. I could also stick it on my to do list, and we can see who gets around to it first.
I tried to post something once and I got the error:
Your post could not be submitted at this time due to the volume of spam and abuse originating from your network.
I was very confused because I do not have a business to promote, and I don't recall ever trolling recently. So if this is abou...
