Font Size: a A A

Anti-spam System And Realization

Posted on:2006-01-03Degree:MasterType:Thesis
Country:ChinaCandidate:B WangFull Text:PDF
GTID:2208360152998489Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
This paper concerns about "the design and realization of the anti-spam system". It makes a deep research of completely new structure of anti-spam filter and its system. Taking research on the system has great applied and theoretical values. The most important application of Internet is E-mail service. There are about 50 percent of spams in 1.4 billion E-mails every day. They eat up a large number of network resources, system resources and store resources of E-mail servers. Because the reasons mentioned above, anti-spam becomes more and more serious problem in computer network communication. With reference to lot of native and foreign materials, we took deep research and analysis on anti-spam system and proposed a new method to solve the problem. The paper includes three parts. First part includes 1st, 2nd, 3rd chapter that introduce the protocols of E-mail system, the definition of spam and the basic technologies of anti-spam system. The second part, including 4th, 5th, 6th, 7th chapter following above content, gives deep research on key module –context filter of anti-spam and bring forward a new filter based on ANN plus GA. By using genetic algorithms, which is good at wide searching ability in solution space on optimizing connection weight matrix of artificial neural network, artificial neural network can get over the inherent limitation of local minimal and improve its learning speed and recognition ability. The application of the model in spam filter has been successfully proved by its implementing. Final part, including 8th, 9th, 10th chapter, takes more concern on deploying of filter and compares the traditional deploying with new one. Analyzing drawbacks of current anti-spam technologies and traditional network deploying of anti-spam mail system, this paper presents a new anti-spam network deploying and its implementing, which is based on mature mobile agent technology. At the end of this paper, the details of design of agent base and mobile agent are also given, followed by analyzing on the new strategy using test result. At the last of the paper, we bring forward the key and difficult issues of anti-spam project and make it certain that mail filter and deploying of network are most important approaches to solve the problem.
Keywords/Search Tags:Spam, Mail Filter, Artificial Neural Network, Genetic Arithmetic, Network Deploying
PDF Full Text Request
Related items