For quite sometime naive bayesian classifier based SPAMBayes filtered my emails very accurately with very few false positives.在相當一段時間的樸素貝葉斯分類基於spambayes過濾我的電子郵件非常準確，很少假陽性。

Recently however I have noticed few trends in spamming which are alarming in nature.最近，但我注意到數的趨勢，垃圾郵件這是令人震驚的性質。

  • Database poisoning: Using otherwise innocuous words (ham words) in a SPAM, thereby effectively poisoning the database in the long run數據庫中毒：使用無害的，否則的話（火腿字）在一個垃圾郵件，從而有效地中毒數據庫，在長遠而言
  • Junk Tags: Hiding spam words by inserting invalid HTML tags in between words.垃圾標籤：隱藏垃圾郵件的話，插入無效的HTML標記，在字與字之間。 Any HTML parser ignores tags it doesn’t understand, thereby resulting in properly viewable document任何HTML解析器忽略了標記，它不明白，因而在適當的檢視文件
  • Invalid Words: Spam word like mortgage etc. are masked by inserting special characters or junk characters in between.無效的話：垃圾郵件字一樣，按揭等都是蒙面插入特殊字符或垃圾的人物之間。

Solutions I could think of:解決方案，我可以認為：

  • Most of the database poisoning email tend to be classified in Not Sure category.大部分的數據庫中毒的電子郵件，往往被歸類在不能確定類別。 I suggest that you delete them instead of classifying them as spam.我建議你刪除他們，而不是歸類為垃圾郵件。 However it still requires that we spend some time for it which is what I don’t like.不過，我們仍然需要我們花一些時間，因為這是我不喜歡。
  • Junk Tags: Add a filter in front of bayesian classifier to eliminate junk tags垃圾標籤：添加過濾器在前面的貝葉斯分類，以消除垃圾標籤
  • Invalid Words: No-exact matching algorithms from Lucene etc. should help.無效的話：不完全匹配算法從Lucene的等，應有助。

I have recently noticed a significant increase in mortgage spams.我最近注意到一個顯著增加，在按揭垃圾郵件。 It should be easy to tackle them by legal means.它應該可以很容易地解決這些問題通過法律手段。

Overall the game is becoming tougher for spam prevention.整體遊戲正在成為更嚴厲的垃圾郵件預防。 A combination of existing techniques are required for any spam filters to remain effective.結合現有的技術所需要的任何垃圾郵件過濾器，以維持有效的。

Looking forward to hear your thoughts.期待著聽到您的想法。