A Survey of Various Machine Learning Algorithms on Email Spamming
One of the greatest accepted methods of communication involves the use of e-mail for personal messages or for
business purpose. One of the considerable concerns of using the email is the problem of e-mail spam. The worst part of the
spam emails is that, these are invading the users beyond their consent and bombarding of these spam mails fills up the whole
email space of the user along with that, the issue of the wasting the network capacity and time consumption in checking and
deleting the spam mails makes it even more concerning issue. Spam is a leading headache that attacks the purpose of
electronic mails. So, there is appropriate substantial to distinguish ham emails from spam emails; many methods have been
proposed for classification of email as spam or ham. Spam filtering is a technique which discovers nonessential, unsolicited,
junk emails such as spam emails, and prevents them from getting into the user’s inbox. With the increasing demand of
removing the spam mails the area has become magnetic to the researchers. The filter classification can be categorized into
two techniques - based on machine learning technique and those based on non-machine learning techniques. Machine
learning techniques include Naïve Bayes, Support Vector Machine, AdaBoost, and Decision Tree etc. whereas Non-
Machine Learning techniques are Black/White List, Signatures, Mail Header Checking etc. This paper intends to present the
Comparative Analysis of performance of various pre-existing classification techniques.
Keywords- Classification, E-mail Threats, Spam Filtering, Efficiency.