Dissertation/Thesis Abstract

A structure based technique for spam detection and email classification
by Desai, Varun, M.S., California State University, Long Beach, 2016, 56; 10017844
Abstract (Summary)

Many techniques are available to combat the spread of unwanted emails and online spams. One popular technique is content-based Bayesian filters. Spammers have found techniques to defeat these filters. A structure-based anti-spam technique uses a different approach to the spam problem by checking for the structure of a message instead of its content. The structure of an email is extracted from the DOM (Document Object Model) of the HTML (Hyper Text Markup Language) in the email. We implemented a tree-based comparison and quadratic weighted level scoring system to find similarities between emails. This method is used for email classification so that similar emails can be grouped together. Upon classification of an email, we compared the domain of the email to the whitelisted domains. If the domains do not match we label the email as a spam. The experimental results showed a high success rate of spam detection and email classification.

Indexing (document details)
Advisor: Aliasgari, Mehrdad
Commitee: Englert, Burkhard, Murgolo, Frank
School: California State University, Long Beach
Department: Computer Engineering and Computer Science
School Location: United States -- California
Source: MAI 55/03M(E), Masters Abstracts International
Subjects: Computer science
Keywords: Bayesian filter, Document object model, Email classification, Phising, Spam detection, Structured based technique
Publication Number: 10017844
ISBN: 978-1-339-50539-8
Copyright © 2021 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy