Design And Implementation Of A Anti-spam System Based On Fuzzy Clustering Algorithm

With the rapid development of the Internet, spam in the world has developed to extremely serious degree, users have to spend a lot of time to deal with spam, the server at the same time also need better hardware configuration to the storage and processing of growing mail, which directly reduce the user’s work efficiency, but also a waste of hardware equipment management the rubbish information, how to effectively filter the spam has become an issue of concern, the issue from the point of view using data mining algorithm for Fuzzy Clustering Fuzzy into spam mail and mail is also calculated method, a message is spam and legitimate mail degree, so as to distinguish spam.This thesis firstly illustrates the current situation of spam, and investigates the correlative techniques dealing with spam as well as summarizes the fuzzy clustering algorithms. After that, the paper introduces Fuzzy Set Theory. In the past research on spam filtering, there are not much based on fuzzy clustering. This paper suggests to use the Fuzzy C-Means(FCM) Clustering algorithm based on membership to filter spams. And then the paper compares the performance of FCM and the traditional technology dealing with spam filtering and comes up with the conclusion. In this thesis, the background for the development of practical applications, the use of software engineering principles and development methods, designed and developed a fuzzy C-means clustering algorithm anti-spam filtering system. First conducted a needs analysis development system to get the system functional requirements, and object-oriented modeling using UML for system design techniques. Then the system into functional modules, the system is divided into pre-processing module, filter module, data module three modules, and then describes the main functions of the system development environment and the concrete realization of the module, the module and the key code. Finally, the system spam filtering system, experiments and analysis, and experimental results were analyzed, in summary, analysis and several other algorithms on the basis of performance comparison, has been based on fuzzy C-means clustering algorithm for anti-e-mail filtering system with high performance.The accuracy of spam filtering is always the hotspot of the research, meanwhile the nodus. The algorithm applied in this paper makes the classification of emails fuzzy and then compares with the accuracy of the traditional spam filtering technology. The results of the experiment show that FCM is feasible in filtering spams.