Created Fediverse (markdown)

Namhyeon Go 2024-03-04 17:26:04 +09:00
parent 4cd8061ca9
commit 5cd47dcb7d

13
Fediverse.md Normal file

@ -0,0 +1,13 @@
## Fediverse
### Spam filtering strategy (Fediverse only)
* [K-Anonymity](https://en.wikipedia.org/wiki/K-anonymity) test - Estimating whether the characters has been arranged by humans. (use [Have I Been Pwned](https://haveibeenpwned.com/Passwords))
* Not CAPTCHA - Image spam containing characters that look very similar to [CAPTCHA](https://en.wikipedia.org/wiki/CAPTCHA). (use [TrueCaptcha](https://truecaptcha.org/))
* VowelRatio10 - In characters arranged by humans, there is a high frequency of [vowels](https://en.wikipedia.org/wiki/Vowel) (aeiou) and [semivowels](https://en.wikipedia.org/wiki/Semivowel) (wy) and vowel-ending patterns included in strings that are 10 characters.
* Palindrome4 - Detect [palindromes](https://en.wikipedia.org/wiki/Palindrome) composed of 4 or more characters
* KnownWords4 - Detect [well-known words](https://github.com/dwyl/english-words) composed of 4 or more characters
* SearchEngine3 - In public search engine, the given string yields more than 2 results. (use [LibreY](https://github.com/Ahwxorg/librey))
* RepeatedNumber3 - Detect a repeated numbers 3 times or more.
* SSL decryption (MITM) when relaying to federated servers.
The strategies were implemented to respond to [the Fediverse Spam Attacks which started on the 15th of February](https://github.com/Mastodon-DE/blocklists/blob/main/spam%2F2024-02-15%2F2024-02-15-spam-mute-list.md).