🚀 go-pugleaf

--------------4893E84302EDB4CEEC9F1FFB
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit

You are invited to attend the following Faculty of Information
Technology seminar:

Speakers:  Dr Alison Anderson                Mr. Mal Corney
                Senior Lecturer                       Senior Researcher
                School of Information Systems Information Security
Research Center
                QUT                                      QUT


Title: Profiling Email Text and Traffic for Forensic Purposes

Date: Monday, 20th November 2000

Time: 12.00pm
Place: Lecture Theatre S301 , Faculty of Information Technology,
                                            Gardens Points Campus,
Queensland University of Technology

Abstract:
The amount of email traffic has increased markedly in recent years as
more companies and institutions rely on it for transacting business and
as the general populous embraces the Internet for personal use and
communication.  With this increase in email traffic, comes an
undesirable increase in the use of email for illegitimate reasons.
Examples of misuse of email include spamming (the distribution of
unsolicited junk mail), sending threats, sending hoaxes and distribution
of computer viruses.  Even though each email message carries with it a
wrapper or envelope containing the sender�s details, these details can
be forged or email messages can be routed through an anonymous mail
server therefore hiding the identity of the original sender.  By using
techniques from the fields of authorship attribution (stylometry),
computational linguistics and machine learning, it is thought that a
technique to profile email text and to classify that email to a
particular author can be established.

The central claim of authorship attribution is that there are
objectively described patterns which make author identification possible
or at least permit forming hypotheses that can be pursued by other
means.
Authorship analysis has been undertaken on various literary works by
finding traits or attributes that are unique for a particular author.
If an authorship profile exists, then the classification of authorship
can be reduced to a pattern matching problem by finding a discriminatory
set of attributes.

To date this research has focused on the search for the discriminatory
set of attributes and has used the Support Vector Machine as the machine
learning classifier.  This seminar will present some of the results of
our work to date on different sets of attributes that have been applied
to text documents and to email messages.



--------------4893E84302EDB4CEEC9F1FFB
Content-Type: text/html; charset=iso-8859-1
Content-Transfer-Encoding: 8bit

<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
You are invited to attend the following Faculty of Information Technology
seminar:
<p>Speakers:&nbsp; Dr Alison Anderson&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Mr. Mal Corney
<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Senior Lecturer&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Senior Researcher
<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
School of Information Systems Information Security Research Center
<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
QUT&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
QUT
<br>&nbsp;
<h3>
Title: Profiling Email Text and Traffic for Forensic Purposes</h3>

<h4>
Date: Monday, 20th November 2000</h4>
Time: 12.00pm
<br>Place: Lecture Theatre S301 , Faculty of Information Technology,
<br>&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
Gardens Points Campus, Queensland University of Technology
<p>Abstract:
<br>The amount of email traffic has increased markedly in recent years
as more companies and institutions rely on it for transacting business
and as the general populous embraces the Internet for personal use and
communication.&nbsp; With this increase in email traffic, comes an undesirable
increase in the use of email for illegitimate reasons.&nbsp; Examples of
misuse of email include spamming (the distribution of unsolicited junk
mail), sending threats, sending hoaxes and distribution of computer viruses.&nbsp;
Even though each email message carries with it a wrapper or envelope containing
the sender�s details, these details can be forged or email messages can
be routed through an anonymous mail server therefore hiding the identity
of the original sender.&nbsp; By using techniques from the fields of authorship
attribution (stylometry), computational linguistics and machine learning,
it is thought that a technique to profile email text and to classify that
email to a particular author can be established.
<p>The central claim of authorship attribution is that there are objectively
described patterns which make author identification possible or at least
permit forming hypotheses that can be pursued by other means.
<br>Authorship analysis has been undertaken on various literary works by
finding traits or attributes that are unique for a particular author.&nbsp;
If an authorship profile exists, then the classification of authorship
can be reduced to a pattern matching problem by finding a discriminatory
set of attributes.
<p>To date this research has focused on the search for the discriminatory
set of attributes and has used the Support Vector Machine as the machine
learning classifier.&nbsp; This seminar will present some of the results
of our work to date on different sets of attributes that have been applied
to text documents and to email messages.
<br>&nbsp;
<br>&nbsp;</html>

--------------4893E84302EDB4CEEC9F1FFB--