For quite sometime naive bayesian classifier based SPAMBayes filtered my emails very accurately with very few false positives.在相当一段时间的朴素贝叶斯分类基于spambayes过滤我的电子邮件非常准确，很少假阳性。

Recently however I have noticed few trends in spamming which are alarming in nature.最近，但我注意到数的趋势，垃圾邮件这是令人震惊的性质。

  • Database poisoning: Using otherwise innocuous words (ham words) in a SPAM, thereby effectively poisoning the database in the long run数据库中毒：使用无害的，否则的话（火腿字）在一个垃圾邮件，从而有效地中毒数据库，在长远而言
  • Junk Tags: Hiding spam words by inserting invalid HTML tags in between words.垃圾标签：隐藏垃圾邮件的话，插入无效的HTML标记，在字与字之间。 Any HTML parser ignores tags it doesn’t understand, thereby resulting in properly viewable document任何HTML解析器忽略了标记，它不明白，因而在适当的检视文件
  • Invalid Words: Spam word like mortgage etc. are masked by inserting special characters or junk characters in between.无效的话：垃圾邮件字一样，按揭等都是蒙面插入特殊字符或垃圾的人物之间。

Solutions I could think of:解决方案，我可以认为：

  • Most of the database poisoning email tend to be classified in Not Sure category.大部分的数据库中毒的电子邮件，往往被归类在不能确定类别。 I suggest that you delete them instead of classifying them as spam.我建议你删除他们，而不是归类为垃圾邮件。 However it still requires that we spend some time for it which is what I don’t like.不过，我们仍然需要我们花一些时间，因为这是我不喜欢。
  • Junk Tags: Add a filter in front of bayesian classifier to eliminate junk tags垃圾标签：添加过滤器在前面的贝叶斯分类，以消除垃圾标签
  • Invalid Words: No-exact matching algorithms from Lucene etc. should help.无效的话：不完全匹配算法从Lucene的等，应有助。

I have recently noticed a significant increase in mortgage spams.我最近注意到一个显着增加，在按揭垃圾邮件。 It should be easy to tackle them by legal means.它应该可以很容易地解决这些问题通过法律手段。

Overall the game is becoming tougher for spam prevention.整体游戏正在成为更严厉的垃圾邮件预防。 A combination of existing techniques are required for any spam filters to remain effective.结合现有的技术所需要的任何垃圾邮件过滤器，以维持有效的。

Looking forward to hear your thoughts.期待着听到您的想法。