A Survey of Data Leakage Detection and Prevention Solutions

97 Pages·2012·1.714 MB·English

by  Asaf Shabtai, Yuval Elovici, Lior Rokach (auth.)

SpringerBriefs in Computer Science
Series Editors
Stan Zdonik
Peng Ning
Shashi Shekhar
Jonathan Katz
Xindong Wu
Lakhmi C. Jain
David Padua
Xuemin Shen
Borko Furht
VS Subrahmanian
For further volumes
http://www.springer.com/series/10028

Asaf Shabtai Yuval Elovici Lior Rokach
(cid:129) (cid:129)
A Survey of Data Leakage
Detection and Prevention
Solutions
Asaf Shabtai Yuval Elovici
Department of Information Systems Department of Information Systems
Engineering Engineering
Ben-Gurion University Telekom Innovation Laboratories
Beer-Sheva, Israel Ben-Gurion University
Beer-Sheva, Israel
Lior Rokach
Department of Information Systems
Engineering
Ben-Gurion University
Beer-Sheva, Israel
ISSN 2191-5768 ISSN 2191-5776 (electronic)
ISBN 978-1-4614-2052-1 ISBN 978-1-4614-2053-8 (eBook)
DOI 10.1007/978-1-4614-2053-8
Springer New York Heidelberg Dordrecht London
Library of Congress Control Number: 2012932272
© The Author(s) 2012
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfi lms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection
with reviews or scholarly analysis or material supplied specifi cally for the purpose of being entered and
executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this
publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s
location, in its current version, and permission for use must always be obtained from Springer. Permissions
for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to
prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specifi c statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of publication,
neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or
omissions that may be made. The publisher makes no warranty, express or implied, with respect to the
material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)
Preface
Information and data leakage pose a serious threat to companies and organizations
as the number of leakage incidents and the cost they infl ict continues to increase.
Whether caused by malicious intent or by an inadvertent mistake, data loss can
diminish a company’s brand, reduce shareholder value, and damage the company’s
goodwill and reputation. Data leakage prevention (DLP) has been studied both in
academic research areas and in practical application domains. This book aims to
provide a structural and comprehensive overview of current research and practical
solutions in the DLP domain. Existing solutions have been grouped into different
categories based on a taxonomy described in the book. The taxonomy presented
characterizes DLP solutions according to various aspects such as leakage source,
data state, leakage channel, deployment scheme, prevention and detection
approaches, and action taken upon leakage. In the commercial section solutions
offered by the leading DLP market players are reviewed based on professional
research reports and material obtained from vendor Web sites. In the academic sec-
tion available academic studies have been clustered into various categories accord-
ing to the nature of the leakage and the protection provided. Next, the main data
leakage scenarios are described, each with the most relevant and applicable solution
or approach that will mitigate and reduce the likelihood or impact of data leakage.
In addition, several case studies of data leakage and data misuse are presented.
Finally, the related research areas of privacy, data anonymization, and secure data
publishing are discussed.
We would like to express our gratitude for all colleagues and graduate students
that generously gave comments on drafts or counsel otherwise. We would like to
express our special thanks to Jennifer Evans, Jennifer Maurer, Courtney Clark, and
the staff members of Springer for their kind cooperation throughout the production
v
vi Preface
of this book. We would like to thank to Prof. Zdonik, S., Prof. Ning, P., Prof.
Shekhar, S., Prof. Katz, J., Prof. Wu, X., Prof. Jain, L.C., Prof. Padua, D., Prof.
Shen, X., Prof. Furht, B. and Prof. Subrahmanian, V. for including our book in their
important series (SpringerBriefs in Computer Science).
Beer-Sheva, Israel Asaf Shabtai
Yuval Elovici
Lior Rokach
Contents
1 Introduction to Information Security .................................................... 1
2 Data Leakage ............................................................................................ 5
3 A Taxonomy of Data Leakage Prevention Solutions............................. 11
3.1 What to protect? (data-state) ............................................................. 11
3.2 Where to protect? (deployment scheme) ........................................... 12
3.3 How to protect? (leakage handling approach) ................................... 13
4 Data Leakage Detection/Prevention Solutions ...................................... 17
4.1 A review of commercial DLP solutions ............................................ 17
4.1.1 Market overview .................................................................... 17
4.1.2 Technological offerings of market leaders ............................ 17
4.1.3 Conclusions, remarks, and problems
with the state of the art in industrial DLP ............................. 19
4.2 Academic research in the DLP domain ............................................. 21
4.2.1 Misuse detection in information retrieval (IR) systems ........ 24
4.2.2 Misuse detection in databases ............................................... 26
4.2.3 Email leakage protection ....................................................... 28
4.2.4 Network/web-based protection ............................................. 30
4.2.5 Encryption and access control ............................................... 31
4.2.6 Hidden data in fi les................................................................ 33
4.2.7 Honeypots for detecting malicious insiders .......................... 34
5 Data Leakage/Misuse Scenarios ............................................................. 39
5.1 Classifi cation of data leakage/misuse scenarios ................................ 39
5.1.1 Where did the leakage occur? ............................................... 39
5.1.2 Who caused the leakage? ...................................................... 39
5.1.3 What was leaked? .................................................................. 40
5.1.4 How was access to the data gained? ...................................... 40
5.1.5 How did the data leak? .......................................................... 41
5.2 Description of main data leakage/misuse scenarios .......................... 41
5.3 Discussion ......................................................................................... 46
vii
viii Contents
6 Privacy, Data Anonymization, and Secure Data Publishing ................ 47
6.1 Introduction to data anonymization ................................................... 47
6.2 Elementary anonymization operations .............................................. 48
6.2.1 Generalization ....................................................................... 48
6.2.2 Suppression ........................................................................... 51
6.2.3 Permutation ........................................................................... 51
6.2.4 Perturbation ........................................................................... 51
6.3 Privacy models .................................................................................. 52
6.3.1 Basic concepts ....................................................................... 52
6.3.2 k-Anonymity ......................................................................... 52
6.3.3 L-Diversity ............................................................................ 54
6.3.4 K-Uncertainty ........................................................................ 55
6.3.5 (X,Y)-Privacy ........................................................................ 56
6.3.6 (X,Y)-Anonymity .................................................................. 56
6.3.7 (X,Y)-Linkability .................................................................. 57
6.4 Metrics ............................................................................................... 58
6.4.1 Information metrics ............................................................... 58
6.4.2 Search metrics ....................................................................... 59
6.5 Standard anonymization algorithms .................................................. 60
6.6 Multiple-release publishing ............................................................... 62
6.6.1 Single vs. multiple-release publishing .................................. 63
6.6.2 Publishing releases in parallel ............................................... 63
6.6.3 Publishing releases in sequence ............................................ 64
6.6.4 Anonymizing sequential releases .......................................... 64
7 Case studies ............................................................................................... 69
7.1 Misuse detection in database systems ............................................... 69
7.1.1 Applying unsupervised context-based analysis .................... 70
7.1.2 Calculating a misusability score for tabular data .................. 74
7.2 Using honeytokens ............................................................................ 76
7.3 Email leakage .................................................................................... 79
8 Future Trends in Data Leakage .............................................................. 83
References ....................................................................................................... 87
Chapter 1
Introduction to Information Security
The NIST Computer Security Handbook [NIST, 1995] defi nes the term c omputer
security as “protection afforded to an automated information system in order to
attain the applicable objectives of preserving the integrity, availability, and confi -
dentiality of information system resources (includes hardware, software, fi rmware,
information/data, and telecommunications).” The security concepts of confi dential-
ity, integrity and availability are also called the CIA triad.
Confi dentiality of information is typically seen as assurance that sensitive infor-
mation is accessed only by authorized users. This task can be achieved by various
mechanisms such as encryption and access control.
Integrity of information is typically seen as assurance that information is not
modifi ed by unauthorized users in a way that authorized users will not be able to
identify the modifi cation. This task can be achieved by various mechanisms such as
digital signatures and message authentication code.
Availability is the task of ensuring that a system provides its services to its users
at any point in time. Usually a system includes many mechanisms to ensure its
availability, such as use of several independent power sources and multiple com-
munication lines.
Nonrepudiation, access control, authentication, and privacy are concepts that are
considered part of computer security as well.
Nonrepudiation ensures that a user who sends a message cannot deny that she is
the originator of the message and furthermore, that the receiver of the message can
prove in a court of law that she received the message from the sender.
Access control is the task of controlling which information and services a user
may access after being identifi ed. An access control mechanism can be used only if
the user has been initially identifi ed by the system. In many systems, every autho-
rized action of the user is recorded by an a udit mechanism.
Authentication is the task of verifying the identity of users who connect to a com-
puterized system. This task can be achieved by the user’s providing a unique secret
A. Shabtai et al., A Survey of Data Leakage Detection and Prevention Solutions, 1
SpringerBriefs in Computer Science, DOI 10.1007/978-1-4614-2053-8_1,
© The Author(s) 2012
2 1 Introduction to Information Security

Organization
impose
wish to
assess the
minimize
value of
that
Security reduces to Organizational
Security risk
mechanisms assets
that may
that may be
possess
reduced by
wish to
leading to
Vulnerabilities abuse
that
to and/or may
increase
damage
that
exploits
Threats Attacker
give rise to
Fig. 1.1 Relationship among security terminology players (adapted from Stallings and Brown
(2007))
such as a password (the user proves “what she knows”), using a unique token that
the user possesses (the user proves “what she has”), or using some kind of biometric
identifi cation mechanism such as fi ngerprint (the user proves “what she is”).
Privacy relates to the task of certifying that a user has control of information col-
lected about her and exposed to others. It is diffi cult to defi ne the specifi c mecha-
nisms used to ensure such user privacy. The entire system must be designed so that
the user’s privacy will not be violated.
Ensuring computer security is an extremely challenging task [Elovici, 2012]. In
many cases, the security requirements are clear; however, it is less clear how to use
the various mechanisms to meet these requirements. The security mechanism in
many cases may become the next sensitive part of the system. For example, forcing
the user to use a complex password may result in the user’s writing a note with the
password attached to the computer screen. Computer security is a continuous battle
between the attackers who identify new security holes and vulnerabilities in systems
and the organization’s security department who must prevent them.
This book uses the security terminology proposed in [Stallings, 2007], which is
presented in Figure 1 .1 . It is described here in the context of data leakage and data
misuse.