Unknown Files And Personal Relationship

Published Date: 02 Nov 2017

Xun Huang, Michail Marinos Dimakogiannis

Department of Mathematics and Computer Science

Eindhoven University of Technology

Abstract

This paper introduces the investigation on how people will do when they receive files from different group of senders who are grouped by the closeness of the relationship. The method what we use is a questionnaire in public places and online, in order to prevent the risk event from happening caused by sending real unknown files to the receivers to investigate. Many aspects have been taken into consideration. At the beginning we will provide some information to the participants before we ask them to fulfill the questionnaire. At the end of the investigation, an analysis is done to find out the regularities or some other characteristics of people from different background and in different degree of closeness.

Key words

Questionnaire; unknown files; closeness; friendship; Malicious

Unknown files and personal relationship

Introduction

Background

Nowadays, more and more people, especially young people, chat with their friends via the internet. Instant messaging software (MSN, Skype etc.) and e-mail are made people's life easier than before, because they give the opportunity to communicate, to share and inform people about what happened around the world. Despite all the benefits, we must be aware of the safety because, through this software, anyone can share dangerous files to the others. (Symantec. (n.d.))

Security attacks on the computers, like worms, phishing attacks and malware attacks are very common these days. Every system is threatened by malicious attacks because of the different kinds of viruses and malwares. (Sharma, V. 2011)

"Suspicious" files are created and uploaded every day and transferred through forums or directly via e-mail as an attach file and/or even through friends/ relatives which send us some of these suspicious files via instant messengers. Actually, most of the times, if some strangers send us an unknown file we will not open it. What will happen when a friend or relative send us an unknown file? Or What if the account of our friend had been hacked? (Which mean that the hacker can steal your personal information, by pretending that he/she is a friend of you [2]).

During the last two decades, The numbers of graphics, audio and video viruses attack via emails and instant messages have increased. From personal experiences and also information that we gather, these kinds of attacks aren’t too complicated to create them. Even an expert, who knows how to protect himself, may not pay attention to take an ordinary-looked on a photo as malicious code. But sometimes, they really are. According to Mikko Hypponen, director of antivirus research for F-Secure, antivirus software will strain to find JPEG malware, because by default, it only searches for .exe files. "Normal antivirus software, by default, will not detect JPEGs," Hypponen said. "You can set your antivirus scanner to look for JPEG, but the trouble is that you can change the file extension on a JPEG to so many things." (Evers, J. 2004)(Llett, D. 2004)

Other kind of attack is the Zip bomb, vulgarly the Zip file contains a nasty payload. "It also has several innocent name. If activated, it will create a randomly named DLL file, several registry keys and restart it through their own e-mail engine to send spam and denial-of-service attacks". (Skarbek, G. 2005)

survey

So we decide to conduct a questionnaire to different kind of persons about what they will do, if they receive a file from their friends, by email, MSN, Skype or something like that.

The purpose of the questionnaire is to investigate to what extend can the closeness of personal relationship exert influence on what people will do after they receive a file from their friends or others.

Method

Overview

We try two methods to deliver the questionnaire. The first method was to have a print form of the questionnaire on papers and be distributed in the library. After testing on some pilots, we found out that this method has low efficiency, and people who are in public places they are mostly busy doing their own works. So we decided to proceed and improve the method by adding a new way to collect the questionnaire data, which was through an online questionnaire sheet. We set up a Google sheet-spread online form for people to fill in. In addition, we send them privately in each participant and we inform them about the survey.

Subject selection

The selection will be randomized, both Asian people and European people or people from other continent. Some of the participants will be students of TU/e which have been chosen randomly by sending email, or by filling the paper form which was the first method, also some friends or relatives in our own countries. Those who received our email will see a hyperlink in their email, and we make it clear that this is a real questionnaire and not a malicious link. We use our TU/e email to send the email to the people that we don’t know. However, in our friends or relatives we use our private e-mails to send the questionnaire. Although, we call some people by cell phone to invite them to join the test because they seldom aware of the email.

There is no difference in questionnaire itself between the people who are answered by paper and who are answered by email.

About ethical issues

And the e-questionnaire will not contain any files witch will destroy the experiment’s data.

Analysis

After finishing our data collection, we will use the SPSS 21 to analyze them. We have collected data from more than 130 people, but for the sake of the simplicity of dealing with the data, we chose the first 120 subjects out of the 130. For every 10 people from 1 to 120, the answer of one question would be used as one condition. (i.e. people number 1 to 10 use their first questions, 11 to 20 use their second questions and so on.) Randomness could be guaranteed in this way to some degree. And the distribution of all the 12 questions to the 120 subjects would be well-proportioned.

First, we checked the frequency table, and the data is correct. So we want to simplify the Majors of the subjects, as both computer science and electronic engineering students are engineering students. The same methods will be used on other majors, so we have a new simplified variable, the ClassifiedMajor.

After that, as we needed to observe what people will do, we observed if the people would open the file they received, so we recoded the variable Open to Open2, the variable Open2 only had two choices, one is 1(open), the other one is 0(other). Because for logistic regression, the dependent value must assume exactly two values on the cases being processed.

From table 3.1, all the correlations are far less than 0.700, so we know that the regression results afterwards will be reliable.

From table 3.2, we can see that the variable Friendship has a Ratio of 3.449, while others’ ratios are around 1 or less than 1, so we can come to a conclusion that "Friendship" is the most significant factor to affect the result of the experiment.

Result

Table 3.1: Correlations

Gender of the participant

The relationship between the people who sent you files and you

The type of the file that the sender sent

If the sender introduced or explained something about the file he sent

Recode the major

Gender of the participant

Pearson Correlation Sig.(2-tailed)N

1

120

-.105

.252

120

.086

.350

120

.086

.350

120

-.314

.000

120

The relationship between the people who sent you files and you

Pearson Correlation Sig.(2-tailed)N

-.105

.252

120

1

120

.000

1.000

120

.000

1.000

120

.220*

.016

120

The type of the file that the sender sent

Pearson Correlation Sig.(2-tailed)N

.086

.350

120

.000

1.000

120

1

120

.000

1.000

120

.043

.645

120

If the sender introduced or explained something about the file he sent

Pearson Correlation Sig.(2-tailed)N

.086

.350

120

.000

1.000

120

.000

1.000

120

1

120

-.099

.281

120

Recode the major

Pearson Correlation Sig.(2-tailed)N

-.314**

.000

120

.220*

.016

120

.043

.645

120

-.099

.281

120

1

120

Table 3.2: Variables in the Equation

95% C.I for EXP(B)

B

S.E.

Wald

dr

Sig.

Exp(B)

Lower

Upper

Step

Introduced

-.862

.461

3.495

1

.062

.422

.171

1.043

ClassfiedMajor

.133

.122

1.191

1

.275

1.142

.899

1.451

FileType

-.938

.464

4.082

1

.043

.391

.158

.972

Friendship

1.238

.313

15.627

1

.000

3.449

1.867

6.372

Gender

.057

.483

.014

1

.906

1.059

.411

2.731

Constant

-1.587

.592

7.182

1

.007

.205

a.Variable(s) entered on step 1: Introduced, ClassifiedMajor, FileType, Friendship, Gender.

Conclusion

After the analyze phase many interesting results came out about the reactions of people when they receive an unknown file. First of all, we notice that when people receive a file from a friend, they are most likely to assign a higher level of trust to the file, if they trust this particular friend as well. Secondly from the table 3.2 we observe that the significances of Gender and ClassfiedMajor are quite similar, because their correlation is medium. For the data we know that most of engineering students are male, while most humanity students are female. Finally, we see also in "introduced or not","FileType" that exert influence on the result, but not too much.

References

[1] Evers, J. (2004). JPEG vulnerability on Windows. Retrieved from http://news.techworld.com/security/2236/jpeg-vulnerability-on-windows/

[2] Free Essays - Computer Science Essays. (n.d.). Internet Security. Retrieved from http://www.ukessays.com/essays/computer-science/internet-security.php

[3] Llett, D. (2004). JPEG exploit could beat antivirus software.. Retrieved from http://news.cnet.com/2100-7349_3-5388633.html

[4] Skarbek, G. (2005). Zip file contains a nasty payload. . Retrieved from http://www.smh.com.au/news/tips/zip-file-contains-a-nasty-payload/2005/07/25/1122143759158.html

[5] Vishrut Sharma (2011). An Analytical Survey of Recent Worm Attacks.

  IJCSNS International Journal of Computer Science and Network Security, 11, 99-103. Retrieved from http://paper.ijcsns.org/07_book/201111/20111116.pdf

[6] Symantec. (n.d.). Symantec Mail Security 8300 Series Technical White Paper.. Retrieved from http://eval.symantec.com/mktginfo/enterprise/white_papers/whitepaper_mail%20_security_8300_series.pdf