Filtering: A few different bots were flagged as "Not Matched"

ThioJoe / YT-Spammer-Purge

Allows you easily scan for and delete scam comments using several methods.

GNU General Public License v3.0

4.57k stars 389 forks source link

Filtering: A few different bots were flagged as "Not Matched" #921

Closed JMVRy closed 1 year ago

JMVRy commented 2 years ago

Filter Mode

Auto-Smart Mode

Select the Problem

Some spammers not detected

(Optional) If 'Other', Enter Very Short Description

No response

Spammer Example / Sample

YTSpammerPurge 2 15 0-Beta1_oThTvUBnBP

Video / Post Link

https://www.youtube.com/watch?v=cElSBdJjUEw

(Optional) Additional Info / Context

After noticing a multitude of spam comments, I ran this tool. The Auto-Smart mode looked the best, so I ran using it. As you can see from the screenshot provided, there were a multitude of bot comments that I believe should've been caught, but weren't for some reason or another. Two of said commenters blatantly wrote the n-word, and yet they were not matched.

blahguy88 commented 2 years ago

To add to this post, here are some links to sample comments:

Firecul commented 2 years ago

The "don't read" comments are caught in the latest version of the program at least. These are the only ones not caught by v2.16.10 on my machine

It is strange those telegram ones make it through though I'll admit. Maybe because it's all one word, it appears to be standard ASCII text Text-me-on-telegram…markiplier000 and text_me_on_telegram...Markiplier000

Scanning with the v2.17-Beta also gives exactly the same result

ThioJoe commented 1 year ago

A lot of these are too unique where it probably makes sense to just be caught by the duplicate filter. As for the telegram ones, they seem to be gone now, but I think they should be caught with the latest filters