[UA-discuss] Spam Filtering

[Asmus Freytag]

On 2/8/2017 9:31 AM, Andrew Sullivan wrote:
> On Wed, Feb 08, 2017 at 04:47:52PM +0000, Andre Schappo wrote:
>> Recently I tested sending email from my DataMail 小山@电邮.在线 to
>> some people on gmail. Mostly they report that my email to them goes
>> into their gmail Spam/Junk folder. I deduce that the gmail spam
>> filtering software is treating chinese/unicode email addresses less
>> favourably than ascii email addresses.
> That seems like a leap.  How many of those people speak Chinese
> natively?
>
> Most spam systems are Bayesian and have been taught to look for
> patterns that match other spam.  Everyone who has taught gmail to
> recognize "Han in the headers == spam" will have to un-teach it that.
> I don't know about you, but since I'm not a native user of Han
> characters I have certainly taught my spam filters such rules of
> thumb.
>
> A
Precisely. As a forum admin, I would cheerfully ban all non-ASCII 
e-mails for
registration, because my user base is not world-wide, but the spammers
are. Only reason I am not doing that today, is that the software doesn't
know about non-ascii e-mails yet :)

A./