INTELLIGENT E-MAIL PERSONALIZATION SYSTEM

• Main Author: SHANMUGASUNDARAM HARIHARAN
• Format: Article
• Language: English
• Published: Taylor's University 2011-02-01
• Series: Journal of Engineering Science and Technology
• Subjects: E-mail; Summarization; Prioritization; Spam; Filtering; Clustering; Text mining; Extraction
• Online Access: http://jestec.taylors.edu.my/Vol%206%20Issue%201%20February%2011/Vol_6_1__050_060_SHANMUGASUNDARAM%20HARIHARAN.pdf

In Internet era E-mail has become the most important mode of communication in every day life. E-mail offers several advantages like secure delivery, speed, cheaper cost, acknowledgement report, transparent service, and distributed environment. As spammers try to induce large amount of spam or unsolicited mails, managing these E-mails’s in an efficient manner requires huge attention. This paper focus on personalizing the E-mail messages after eliminating the spam messages. The basic step starts with pre-processing the documents and classifying the contents into several folders or categories. The next step is to cluster the documents based on the relativeness they have using cosine similarity metric. This clustering approach is carried out using unsupervised method. The mail messages are the parsed through a filter that would identify the spam immediately. Studies on personalization of mails after spam identification, prioritizing the E-mail’s based on the importance and summarization of were also proposed. The results were quiet promising leading to efficient spam identification providing a platform for further improvements to build a domain independent personalizer system.