clustering and classification of email contents参考文献.pdf

: .
Journal of King Saud University osting by Elsevier . on behalf of King Saud University. This is
an open access article under the CC BY-NC-ND license (-nc-nd//).
1. Introduction emails that are not intended for specific receiver and that are
sent for either marketing purposes, or for scam, hoaxes, etc.
Emails are used by most humans on earth. It is estimated that It is estimated that in 2009 more than 97% of emails were clas-
there are more than 3 billion email accounts of almost half of sified as spam (Elements of Computer Security, 2010). This is
the world population. They are expected to reach 4 billion by why many research papers which studied or analyzed emails
the year 2015 (Email Statistics Report, 2011). Even kids are focused on this aspect (. the classification of emails into
allowed under certain conditions to have email accounts super- spam or not). However, the struggle between spammers and
vised by parents. spam detection tools is continuous where each side is trying
Spam in emails is one of the most complex problems in to create new ways to overcome the techniques developed by
email services. Spam emails are those unwanted, unsolicited the other.
Some local papers that conducted spam assessment (.
Abdullah Al-Kadhi, 2011 paper) showed that the problem is
serious. Authors conducted surveys to assess the current status
* Corresponding author.
of Spam distributio