Thursday, October 18, 2007

Spammers respond!

It's amusing when I see that spammers actually respond to my Chainki anti-spam drive. Following is taken from a new page of spam where clearly a spammer has noticed that I have banned some Chinese characters (see my earlier post, Goodbye Chinese spam), and just rambles about what to do. The page then finishes with masses of links to his Chinese language school. Don't worry if you don't understand what the spammer wrote, but here it is anyway:

Ah ... yes. I assumed that this thread couldn't get approval. Then, after searching and seeing the other thread and with the information given in it, I've browsed my dictionary a little and thought that 搦 is more suitable choice and wrote there.

This is the story of this topic and my response in that topic.

Anyway ... How can we write "here you are"?


It sounds as "no" in the forth tone, meaning "here you are". But, I can't find it in my dictionary. How can we write it in Chinese characters? It is in the Pimsleur Mandarin I lessons.

Thanks ...


Haven't you already posted in this thread? -> What is the character for "Here you are" used in Pimsleur

It is 喏, pronounced nuò (not nòu).

Dictionary ->

n mj
Or have I missed something? __________________ 落 了 片 白 茫 茫 大 地 真 乾 淨 (blogger) 奼 紫 嫣 紅 開 遍 (windows lives spaces) Mesuper, thanks.

Monday, October 15, 2007

Goodbye Chinese spam!

In trying to get rid of spam on Chainki, an important part of the battle is to try and ensure that it does not come back again. And ideally this should be done automatically. Even with a spam filter, if you see that a particular site, like xyz.com is adding spam and you block all entries by that site, it's possible that next day another site, abcdefg.com will spam. A lifetime job to add all the possible spammers!

But I've just clicked one neat way to get rid of Chinese spam - block individual Chinese characters! As parts of the site, such as the French Chainki are not allowed anything but French links (and generally French descriptions), it is clearly spam whenever anybody tries to enter Chinese characters into an edit.

Soooo...I have started to add simply random Chinese characters to the spam filter! The funny thing is that I have absolutely no idea what any of the characters I am banning mean, but it is (I think) likely to be highly effective in stopping Chinese spam. Here is a list of characters I have banned so far, all of which have been copy-pasted completely at random from Chinese spam text I have found on my site (apologies if any of these are swear words!!):














One of the great problems in dealing with spam is that you want to eliminate all the spam of the bad guys easily, without causing any problems to the good guys. This technique I think - or I hope - will eliminate very quickly all the Chinese spam!

Chinese spam? - Bring it on!

Friday, October 12, 2007

Spammers! Grrrr!!!

Just when I thought I had the Chainki spammers on the run, they came back where and when I wasn't looking. I thought, over-confidently that the spammers were finished off by my nifty spam filter, but it does at least need keeping an eye on. I thought the site was largely able to look after itself. I stopped looking at changes.

Mistake!

While I wasn't looking for several months, Chinese spammers put hundreds of pages on the French pages of the site, before I noticed. Now, on the French part of the site, I filter (amongst other keywords):

  • .org.cn
  • .com.cn
  • .net.cn
  • .cn/

which should filter out ALL sites with a Chinese domain name (which have no place on the French pages!). Though that alone doesn't deal with the Chinese sites which use, for example, a dot com domain name, so in addition, I've added a number of particular specific offenders to my spam filter. Time will tell if I need to add many others, but for the moment things seem to be much better.

Another thing I have done is on some parts of the site, made it so that people have to create a login if they want to edit a page. This is a bit of a pity for the casual user who wants to make an improvement, but should also help reduce the casual spammer.

The spam war is one I certainly plan to win, but I do need to keep vigilant!