Content Filtering by Language (specifically Japanese)
Question asked by Brantz - 10/15/2026 at 10:39 AM
Unanswered
I'm getting a lot of Japanese spam from a wide variety of domains and addresses. I'd like to be able to content filter these based on the presence of Japanese characters, in my case the text contains normal Japanese characters mixed with zero‑width characters and special Unicode variants. Is there a way to do this? Thanks.
terry fairbrother Replied
I use routing rules, pick random characters and action = delete / quarantine

this is some of mine

プ
至
限
送
ま

no idea what they mean but appeared the most. I do this with non English words too and really cuts out a lot of spam
 
Brantz Replied
Thanks - that inspired me to take 3 different emails and save them as 3 different files that I then uploaded to Claude and asked it to compare them and only give back the characters common to all 3 files. I then made a content filter to route any messages containing any of those characters in the body to Junk Mail.  Here's what it found:

Visible characters (44)

  • Punctuation and symbols: the space, 、 。 ・ ー
  • Hiragana (28): あ い お が き こ さ し す た て で と な に の は ま よ り る れ を
  • Katakana (2): イ ン
  • Kanji (13): 合 場 定 年 意 手 日 期 本 注 続 記 過
So far I only used the Hiragana and Kanji characters. I'll report back how well this works in a while.

Thanks again!

Reply to Thread

Enter the verification text