| Summary: | 碩士 === 國立臺灣科技大學 === 資訊工程系 === 99 === With the rapid development of the Internet, spam mail has become a business and personal data placement in an important challenge. In addition to traditional commercial spam mail, other attacks, including phishing, pornographic messages, malicious code (viruses) are spread through spam, junk e-mail in addition to caused by a large number of network resource consumption, the more businesses and individuals with to the risk of data leakage.
In this study, we set up a personalized spam filter structure, except by creating a personal blacklist, white list for spam list to determine the outside, and provide users with a feedback mechanism by the user to decide the normal mail or spam e-mail, and e-mail for personal history (including normal mail and spam), via Smooth Support Vector Machine classification of learning, resulting in classification model, to provide follow-up with a message of classification judgments.
|
|---|
