Anti-Spam

A spam filter implemented with Bayes Method. Library

• Mysql connection tool connect database via JDBC in Anti-Spa/lib
• jcseg segments Chinese words
• Tomcat on which this system runs

Bayes Method

• Bayes Method in Spam Detection : In this formula , S = "a spam", w = "a specific word showing in the mail" , N = "a normal email".

  

• An email certainly contains more than one word :

• Consider comprehensively, calculate combining probability( Here we assume that are independent) :

• Further more,we assume , so the formula above can be simplified as:

• I'm using this formula to detect emails in the program.

Here are more details of spam filtering ,fundamental knowledge of Combining Probability and Bayes theorem.

How To Use

• You need firstly include all the libs above to compile this program.Yet if you only uses in Tomcat, mysql connection tool and jcseg are already included in /Anti-Spam/WEB-INF/lib path on Tomcat.

• webapps files of Tomcat are in '/webapps',copy '/webapps/Anti-Spam' folder to $TomcatPath/webapps/ of your Tomcat Installation Path.(attention, /webapps may not be the latest version)

• Before you use it,create a file named 'sqlInfo.ini'(file name defined in org.bit.util.GlobalConstants.java),in which all init infomation is defined.Here are some necessary items.

    url = jdbc:mysql://localhost:3306/test
    user = langley
    password = 123456
    jcsegPropPath = /Anti-Spam/WEB-INF/lib/jcseg.properties #or use absolute path
  

• Run in Tomcat.