Automatic email classification system / Phang Siew Ting
The growing problem of unsolicited bulk email and the growth of the volume of email received has generated a reliable need for email tool that assists the users manage the emails in an effective way. For this purpose, several Machine Learning algorithms has been purposed to automate the classificati...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2003
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/11548/ http://studentsrepo.um.edu.my/11548/1/automatic_email_classification.pdf |
| Summary: | The growing problem of unsolicited bulk email and the growth of the volume of email received has generated a reliable need for email tool that assists the users manage the emails in an effective way. For this purpose, several Machine Learning algorithms has been purposed to automate the classification of emails. The algorithms learn to classify emails based on it textual contents, and subsequently assign individual emails into a predefined set of categories or bins in accordance with the preferences of a user. Automatic Email Classification System is an email reader tool that implements machine learning algorithm in email classification, manipulated by a Graphical User Interface. The interface provided allows building a classifier algorithm, testing it, and applying it to previously unread messages. New messages can be classified and stored into corresponding predefined folder on the fly automatically as they are downloaded from a server. The automated email classification of the system foreseen a greatest advantage over the existing email clients in the market, such as Microsoft Outlook, Netscape Messenger, etc, which rely mostly on hand-constructed keyword-matching rules. |
|---|