clemensv / twitter_spamlists Goto Github PK

IMPORTANT: This is work I am doing for myself because I'm tired of the spam. I am in no way suggesting or recommending for anyone to use these lists. They work for me, you may find them useful.

This work is shared under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0).

The list in this folder has been compiled using a search spider that finds porn spam bots spamming city name hashtags, like #Mönchengladbach or #Viersen as well as "like-spam" from that same group.

The strategy I apply uses a Computer Vision service to verify that included accounts are posting pictures that are "racy", in addition to at least one further criterion. The Computer Vision step is skipped if the picture has already been classified or if the profile shows a pattern known from a spammer group. The nature of the chosen algorithm will occasionally find candidates that are not part of the targeted bot farm, but that nevertheless meet the blocking criteria. Several accounts will look like false positives at first sight, but at closer inspecting you will find that most of them have timelines nearly completely consisting of retweets, which this algorithm ignores. It will rather only look at tweets originating from that account and those will then have been classified as meeting the criteria.

The format is compatible with Twitter's block-list import/export format and uses UserIDs. To see the respective profile, use https://twitter.com/intent/user?user_id={id} for a listed ID.

Please mind that the Twitter import function for block lists is VERY picky and will expect LF line endings and that the last line must not be blank. The ZIP file contains a version of the block lists that will definitely work for import.

Get the ZIP file directly from this link, unpack the file on your PC, and then go to your block list on Twitter and choose the Options drop-down on the right hand side above the list to import.

Twitter has size limits that require me to split this up.

Die Liste in diesem Verzeichnis ist von einem automatischen Suchprogramm erstellt worden, das "Porno" Bots, die Städte-Hashtages wie #Mönchengladback oder #Viersen vollspammen automatisch findet.

Das Format ist kompatibel mit dem import/export Format von Twitter und verwendet numerische Benutzer-IDs. Um das jeweilige Profil zu sehen, verwende https://twitter.com/intent/user?user_id={id} mit einer gelisteten ID.

Die Twitter Import/Export Funktion is sehr anspruchsvoll in Sachen Formatierung der Zeilen- und Dateinenden. Die ZIP Datei enthält Dateien, die sicher funktionieren.

Lade die ZIP Datei direkt von diesem Link, herunter, packe sie aus, und dann gehe zur Blockierliste auf Twitter und wähle "Erweiterte Optionen" auf der rechten Seite über der Liste für den Import.

Die Größenbeschränkungen von Twitter für diese Dateien erfordert, dass ich mehrere Dateien zur Verfügung stelle.

Bitteschön :)