Volume 50, Number 4-5, October-December 2016
|Page(s)||951 - 963|
|Published online||03 November 2016|
Clustering of optimized data for email forensics
1 University Abderrahmane Mira of bejaia, LIMED Laboratory,
06000 Bejaia, Algeria.
2 University college Dublin, Parallel Comptaional Research Group Laboratory, Dublin4, Dublin, Ireland.
Accepted: 12 November 2015
Forensics is a study of evidence to help the police solving crimes. If we apply (Forensics) in Computer Sciences domain, crimes are mainly network attacks found more in emails; which become nowadays the most popular way of communication accessible via Internet. We receive in our Inboxes emails gangs without being aware of them. Therefore, it is necessary to build an automatic checking system to filter good emails from bad ones. In this paper, we propose a new emails processing approach using Singular Value Decomposition method (SVD) to optimize emails data before applying Data Mining techniques (Clustering) to extract bad emails located in the mail servers where the user’s inboxes are hosted. Our study is based on filtering Emails (bads and goods) by the clustering of optimized data compared with unoptimized one.
Mathematics Subject Classification: 05C12 / 05C50 / 05B10 / 91C20 / 15A18 / 34A05
Key words: Email / feronsics / spam / SVD / LSI / optimisation / data mining / clustering
© EDP Sciences, ROADEF, SMAI 2016
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.