Enhancing associative classification technique for spam detection

Dissertant

Jarrah, Muhammad Ahmad Muhammad

Thesis advisor

al-Zubaydi, Rashid

University

Philadelphia University

Faculty

Faculty of Information Technology

Department

Department of Computer Science

University Country

Jordan

Degree

Master

Degree Date

2017

English Abstract

Data mining techniques can extract meaning from noisy data, find out patterns in random data, and use this information to best understand trends, patterns, connection and relations between the data.

Data mining has two classes (supervised and unsupervised) depending on the human management or predefining the goal.

One of the hybrid techniques which combine supervised and unsupervised technique to achieve a full automated process to find hidden patterns and classify the data depending on this patterns and relations is the Associative classification AC.This technique is widely used in the world to solve many problems.

In this study we propose a new enhanced associative classification approach work to increase the accuracy of it by decreasing the number of inaccurate rules that may obtained by the association rule mining by adding a second rule checking step to solve this problem and proved that by the experimental results.

The proposed approach applied in the cyber security domain exactly to solve the SPAM emails problem as a SPAM detection hybrid algorithm (join link based (header) and content base).

The implementation of this approach shows that the proposed approach improves the performance of the traditional approach in different aspects .

For example the precision measure increased from 0.543 to 0.862 and the accuracy measure also increased from 0.721 to 0.890compared with the traditional approach.

Main Subjects

Information Technology and Computer Science

No. of Pages

51

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-955865