Research Of Anti-Spam Based On Mail Purpose And Finger-Print

With the rapid growth of internet, email has become one of the most popular communication tool and being widely accepted by personal and business users. Spam, unsolicited, undesired, bulk email (or junk email), has been a significant security issue for internet users and a massive waste of time, disk space and network bandwidths. Especially today, with the increasing dedication and sophistication of spammers, even popular social media sites such as Tencent (QQ), Sina Weibo, and Google Plus are not exempt from email spam as they all interface with email systems. The explosive growth of unsolicited emails has prompted the development of numerous spam filtering techniques. With an "arms-race" between spammers and spam filter developers, spam has been continually changing over the years. Although spam emails have been transitory reduced after new filtering technology being deployed, spammers are creating new ways against anti-spam technology. This paper includes the following parts:●Take a general study on current spam detection and filtering technology, compare and summarize their advantages and disadvantages, so that we can grasp the latest trend of spam detection and filtering technology;●After a deeply compare and anlyze on mass of spam and ham, we find out the difference on the purpose of the sender, further more we propose a spam filtering system that is base on intention filtering which is tested on a spam archive dateset and is found to be highly effcient and exactly;●Find out the characteristic of image-based spam with text-based spam out of a large amount of emails, summarize the weakness of machine learning based image spam filtering technology, we propose an image spam filtering system which is based on statistical probability, ruling and voting mechanism;●Build up an effcient Hash arithmetic, and caculate the hash-value of spam text and attachment, generate the finger-mark file and which is used to compare with online finger-mark datebase to further tell whether the mail is spam or ham.