Design And Implement Of Intelligent Detecting System Of Phishing Sites On Clients

"Phishing" website attack is by mimicking real sites, to achieve impersonate any user identity fraud, and then steal personal account information or privacy violations. With more and more serious damage of phishing sites, phishing detection as an anti-phishing measures and technologies are subject to widespread concern and attention.This paper presents a SVM classifier machine learning algorithm from phishing intelligent detection methods with feature extraction of Web content, based on the detection and identification with the SVM model classification phishing. First, because the feature selection of web content extraction of feature and phishing large extent influence the detection result, this article fully extracted the various aspect of the web page, and then further demonstrate that the specific characteristics to distinguish between the false and the true Web sites to improve the recognition rate of the fake website. Followed by extraction of web page characteristic is a very time consuming task, in order to further reduce the extraction of characteristic values of CPU time consumed, the idea of taking multi-threaded concurrency is necessary, therefore a detailed description based on multithreaded web crawler design principles and methods is employed. Finally SVM classifier trained a detection model, and thus to identify the fake website.