Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli

With the latest technology and of mail server it exploits the potential of spamming activities to be increase accordingly. People doing spam on they own ways and for their own reasons. Spam is not a virus and it does not contain viruses. Marketing agencies using automated spam structure to perform m...

Full description

Saved in:

Bibliographic Details
Main Author:	Ramli, Ahmad Kamal
Format:	Thesis
Language:	English
Published:	2008
Subjects:	Algorithms Spam filtering
Online Access:	https://ir.uitm.edu.my/id/eprint/64608/1/64608.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my-uitm-ir.64608
record_format	uketd_dc
spelling	my-uitm-ir.646082023-04-19T02:16:00Z Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli 2008 Ramli, Ahmad Kamal Algorithms Spam filtering With the latest technology and of mail server it exploits the potential of spamming activities to be increase accordingly. People doing spam on they own ways and for their own reasons. Spam is not a virus and it does not contain viruses. Marketing agencies using automated spam structure to perform mass mailing on their promotion and marketing events. At present they are few solutions for tackle spam, using commercial, open source and third party organizations to filter the incoming and outgoing messages. By comparing commercial, Bayesian and N-gram algorithm in rejecting spam and predicting ham messages, it will be a very significant research project for choosing the rights tool to minimize spam activities. By using standard series of text messages which consists of spam and ham words, N-Gram algorithm performed very well. It has the ability to predicting the next alphabet and this is much different with Bayesian algorithm. By applying N-Gram in commercial products, user may receive lots of ham messages inside their inbox. From the test itself Bayesian able to detect only 66.66 % accuracy of ham words inside series of messages. However ,N-Gram score 100% for the same exercise and the algorithm itself have the capability to increase the potential of ham or spam weightage. 2008 Thesis https://ir.uitm.edu.my/id/eprint/64608/ https://ir.uitm.edu.my/id/eprint/64608/1/64608.pdf text en public masters Universiti Teknologi MARA (UiTM) Faculty of Computer and Mathematical Sciences Mamat, Kamaruddin
institution	Universiti Teknologi MARA
collection	UiTM Institutional Repository
language	English
advisor	Mamat, Kamaruddin
topic	Algorithms Spam filtering
spellingShingle	Algorithms Spam filtering Ramli, Ahmad Kamal Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli
description	With the latest technology and of mail server it exploits the potential of spamming activities to be increase accordingly. People doing spam on they own ways and for their own reasons. Spam is not a virus and it does not contain viruses. Marketing agencies using automated spam structure to perform mass mailing on their promotion and marketing events. At present they are few solutions for tackle spam, using commercial, open source and third party organizations to filter the incoming and outgoing messages. By comparing commercial, Bayesian and N-gram algorithm in rejecting spam and predicting ham messages, it will be a very significant research project for choosing the rights tool to minimize spam activities. By using standard series of text messages which consists of spam and ham words, N-Gram algorithm performed very well. It has the ability to predicting the next alphabet and this is much different with Bayesian algorithm. By applying N-Gram in commercial products, user may receive lots of ham messages inside their inbox. From the test itself Bayesian able to detect only 66.66 % accuracy of ham words inside series of messages. However ,N-Gram score 100% for the same exercise and the algorithm itself have the capability to increase the potential of ham or spam weightage.
format	Thesis
qualification_level	Master's degree
author	Ramli, Ahmad Kamal
author_facet	Ramli, Ahmad Kamal
author_sort	Ramli, Ahmad Kamal
title	Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli
title_short	Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli
title_full	Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli
title_fullStr	Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli
title_full_unstemmed	Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli
title_sort	experimental analysis on the anti spam effectiveness - commercial, bayesian and ngram algortihm / ahmad kamal ramli
granting_institution	Universiti Teknologi MARA (UiTM)
granting_department	Faculty of Computer and Mathematical Sciences
publishDate	2008
url	https://ir.uitm.edu.my/id/eprint/64608/1/64608.pdf
_version_	1783735498284466176

Experimental analysis on the anti spam effectiveness - commercial, Bayesian and Ngram algortihm / Ahmad Kamal Ramli

Similar Items