@Makyen Matched by ^[a-z]{4,}erk(?<!Landoperk)(?<!codeherk)(?<!waltwhiterk)(?<!SilverK)(?<!NacOJerk)(?<!Stromwerk)(?<!SpencerK)(?<!Bleuderk)(?<!GameMasterK)(?<!FrazzerK)$(?#Add additional FP usernames when detected; original: 1547532983 iBug) on line 7991 of watched_keywords.txt
@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.
@quartata Basically, it's "link at start of post", with multiple domains excluded and excluding posts that contain <code> or <blockquote>. Its intent has been to collect data prior to moving it to it's own detection reason, where it's more obvious what it's doing and significantly easier to exclude a list of domains. That should have happened a while ago, but I'm running behind. The last I checked, it was running about 59% TP, with multiple posts where it's the only detection to catch TPs.
@Makyen That's 59% TP the last time I checked, based only on posts detected since the watch was added (actually since the original version was added, which was broader and had more FP). Looking at all posts in MS, i.e. including those from before it was added, shows a %TP in the mid to high 90's, which just highlights the biased nature of the data that MS has.
About a month later, and I'm still behind in moving it into its own detection. sigh
@tripleee 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.
@tripleee 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.
[ SmokeDetector | MS ] Bad keyword in body, pattern-matching product name in body, potentially bad keyword in body, potentially problematic ns configuration in body (196): Kam Kashariya Capsules by user327960 on apple.SE
@tripleee Yeah, it could be done by someone else, but I have intended to do it. There's a bunch of combing through the results to find other sites which should be excluded and then creating a new detection.
[ SmokeDetector | MS ] Bad keyword in body, blacklisted website in body, link at end of body, pattern-matching website in body, potentially bad ip for hostname in body (491): LaPura Cream reviews by Johxdwards on meta.SE
[ SmokeDetector | MS ] Bad keyword in body, blacklisted website in body, blacklisted website in title, pattern-matching product name in body, potentially bad keyword in body, +1 more (395): www.townscript.com/e/miraculoux-keto/ 123 by lgyar2ch on stackoverflow.com
[ SmokeDetector | MS ] Bad keyword in body, blacklisted website in body, potentially bad ns for domain in body, potentially bad keyword in body (197): G Streamer Error by Viante934 on askubuntu.com
[ SmokeDetector | MS ] Potentially bad ns for domain in body, potentially bad keyword in body, potentially bad keyword in title (3): Quick Results Keto by ohisamic on askubuntu.com
[ SmokeDetector | MS ] Bad keyword in body, bad keyword in title, blacklisted website in body, body starts with title and ends in url, pattern-matching product name in body, +1 more (490): Helix Male Enhancement by RudyMeser on wordpress.SE
[ SmokeDetector | MS ] Bad keyword in answer, potentially bad asn for hostname in answer, potentially bad ns for domain in answer, potentially problematic ns configuration in answer (89): XPT to CSV Conversion? by Chris on stackoverflow.com
@Machavity this wasn't weight's fault. Some people select in their conditions that they want to autoflag every post which has 3 reasons caught regardless of the reason weight.
@Machavity Yeah, that one is a bit problematic. The addition of more detection reasons to SD significantly skews using just number of detection reasons as a flagging criteria. IIRC, Stack Overflow is particularly problematic. However, 4 reasons is usually quite a good indicator (45,518 TP and 8 FP, which is 99.98%).
@SmokeDetector The code provided appears to have nothing to do with the question. It's just code intended to display the syllabus contents from the user's site. It does not appear to use the tech (blogger) which is asked about in the question.
@SmokeDetector The user's profile indicates the user is associated with the site being promoted.
@double-beep English Language & Usage and The Workplace only require 3 flags to nuke a post. Thus, autoflags are limited to lower numbers than on other sites.
@Chris if you mean in metasmoke, nobody's sent any feedback on it yet. If you mean why hasn't it been deleted/locked by spam flags, we don't have the power to do that unilaterally, only by enough people spam flagging it.
@ArtOfCode Ah, ok. I was just wondering, usually any post which gets spam flagged gets a number of spam flags within minutes. Here on both postings none was raised.
