CLASSIFICATION TECHNIQUES USING SPAM FILTERING EMAIL

Main Article Content

P. PRIYATHARSINI
Dr. C. Chandrasekar

Abstract

The general data mining model with the complex sample data solves the problem on data classification. The preprocessing step of complex data in data mining solves the problem of accuracy caused by the mass data.
The growing volume of spam mails annoys people and affects work efficiency significantly. The work focused on developing spam filtering algorithm, using statistics or data mining approach to develop precise spam rules. The main propose of an anti spam approach combining both data mining and statistical test approach. The efficiency of spam rules, only significant rules will be used to classify emails and the rest of rules can be eliminated for performance improvement.
The effective decision tree classifiers are used to classify whether the mail is spam or ham. Various filtering techniques are used to find the spam mails and filter them but the accuracy and performance of the algorithms is distinct from each other. Two decision tree algorithms that are basically used as classifiers namely J48 or C4.5, Rndtree. The algorithms are studied, analyzed and test results are shown in WEKA tool for efficient spam filtering.The results are compared and RndTree algorithm shows almost 99% accuracy level in filtering the spam mails and it shows best results among other classifiers.

Downloads

Download data is not yet available.

Article Details

Section
Articles