ONTOLOGY EDITING TOOL FOR INFORMATION SECURITY

AND HUMAN FACTORS EXPERTS

John C. Mace, Simon Parkin and Aad van Moorsel

School of Computing Science & Centre for Cybercrime and Computing Security (CCCS)

Newcastle University, Newcastle, U.K.

Keywords: Information Security Ontology, Ontology Editor, Human Factors, OWL.

Abstract: Currently ontology development is facilitated by generic ontology editing tools which accommodate

ontology experts, and not necessarily those individuals whose knowledge requires capture. Furthermore the

process of knowledge capture is time consuming, error prone and requires appropriate technical skills. We

propose a graphical editing tool for ontology development which simplifies knowledge capture. To

demonstrate the tool’s potential this paper explores the need for information security and human factors

experts to capture their related knowledge in a dedicated ontology. Population of such an ontology would

provide supporting information to enable organisations to make more well informed security policy

decisions. Tailoring the tool for information security and human factors experts allows them to produce

ontology content without being familiar with ontology construction and technologies. The tool is intuitive,

requires no ontology component configuration, and provides mechanisms to guide users and reduce the

potential for errors. Our tool allows information security domain experts to develop and extend the

ontology, and organisations to tailor ontology content to their own requirements.

1 INTRODUCTION

There are instances where use of knowledge benefits

from encoding the knowledge in an ontology.

However, those in possession of the knowledge that

is to be captured may be unable to familiarise

themselves with ontology editing tools. This may be

because they do not have the appropriate technical

skills, or the time or inclination to understand

ontology technologies.

Here we present an ontology editing tool that

specifically serves a particular knowledge domain,

in such a way that it negates the need for knowledge

owners to understand ontology construction and

ontology languages. Users construct a graphical

representation of an ontology which can be

automatically translated into an OWL (Web

Ontology Language) encoded ontology.

The approach is demonstrated by applying the

tool to serve information security practitioners.

Within organisations it is essential to consider not

only external security guidelines and standards (e.g.

ISO27001 (BSI27001, 2005)), but also the impact of

security upon people and processes within the

organisation. If CISOs were able to augment their

own expertise of IT security with knowledge of

human-behavioural factors in information security

they would potentially make more informed security

policy decisions.

Applying this tool to information security policy-

making demonstrates a solution where there is a

need to formalise disparate, interconnected domain

knowledge supplied by individuals who require

assistance in the knowledge-capture process. It

cannot be assumed that these individuals are experts

in ontology development. We consider how Chief

Information Security Officers (CISOs) and human

factors experts would create and/or modify an

information security ontology that could

subsequently be tailored to specific organisations.

The tool addresses the need that these individuals

have for a direct and intuitive way to populate an

information security ontology with their

interconnected knowledge of security issues and

human-behavioural factors.

The rest of the paper is arranged as follows:

Section 2 discusses the need for such an ontology

and current construction approaches, alongside

related work in ontology tools and construction.

207

Mace J., Parkin S. and van Moorsel A..

ONTOLOGY EDITING TOOL FOR INFORMATION SECURITY AND HUMAN FACTORS EXPERTS.

DOI: 10.5220/0003092302070212

In Proceedings of the International Conference on Knowledge Management and Information Sharing (KMIS-2010), pages 207-212

ISBN: 978-989-8425-30-0

 2010 SCITEPRESS (Science and Technology Publications, Lda.)

Section 3 discusses the implementation of the editor

tool. Concluding remarks are presented in Section 4.

2 BACKGROUND

2.1 Need for an Information Security

Ontology

Currently, information security policy management

decisions within an organisation are driven by

external security standards and the experience of the

Chief Information Security Officer (CISO) or

similar.

However, policy decisions may affect or be

influenced by the behaviour of individuals within the

organisation. To make policy decision-making more

effective, organisations (and more so CISOs) must

have an awareness of the usability and other human-

behavioural factors inherent in these decisions

(Skidmore, 2003).

There is a need to organise and standardise

disparate IT security policy-making information and

human factors concerns in the form of a knowledge

base or ontology. This would facilitate clear and

effective communication within the IT-security

management community, thereby further informing

the security policy decision-making process.

2.2 Current Ontology Development

Currently the construction and/or modification of a

security ontology is achieved through an ontology

editing tool. These may come in the form of

graphical or textual editors. Both types of editor

allow content to be converted to a file written in an

ontology language.

Both of these types of editor place similar

demands on the user. The user must define the

overall structure for potential ontology content,

including concept types (ontology classes) and

relationship types (ontology properties). As content

is entered, each concept (ontology individuals) must

be individually defined along with its class type,

properties and relationship to other concepts.

This complex process assumes familiarity with

ontology technologies. As such a CISO or Human

Factors Researcher (HFR) may be unable to develop

ontology content themselves, and would require

either the assistance of an ontology expert or a

dedicated ontology editing tool that hides ontology

complexity.

2.3 Related Work

A large amount of work has been carried out in the

field of security ontologies and with the rising

interest in the Semantic Web this work is

supplemented with a vast array of ontology creation

tools.

The capture of security knowledge in an

ontology has been shown to be viable through a

number of studies e.g. (Parkin, 2009) and (Fenz,

2007). The work of Fenz et al (Fenz, 2007)

incorporates the ISO27001 guideline (BSI27001,

2005) with a security ontology that considers the

physical aspects of IT security management. This

work allows organisations to audit security polices

and assess whether they adhere to the ISO27001

guidelines.

For the successful development of a security

ontology, the use of an ontology editing tool is

required. A number of tools are already available

e.g. Protege 3.4.4 (Stanford, 2010), OntoStudio

(Ontoprise, 2010), TopBraid Composer

(TopQuadrant, 2010) and NeOn Toolkit (NeOn,

2010). Protégé 3.4.4 offers form-based content

entry, with ontology content presented in textual

format. Ontology content is organised into class,

property and individual hierarchies, in a manner

whose level of complexity is appropriate for an

ontology expert.

Another tool, SWOOP, (MINDSWAP, 2004)

offers a Web browser style user interface aimed at

the ‘average Web user’ to facilitate the easy

development and browsing of OWL ontologies.

Ontology navigation and editing is carried out via a

hyperlink based system. This approach enables both

ontology and domain experts to contribute but is still

reliant on ontology experts to contribute to the

underlying formal structure for that knowledge.

Visualisation of an ontology during its

construction or modification is of great advantage to

the user and eases these processes immensely. There

are a number of visual ontology creation tools using

OWL as a base language e.g. GrOWL (Vermont,

2006), OWL-S Editor (Scicluna, 2004) and

SemanticWorks (Altova, 2010) all of which

illustrate the ontology in a UML format. SemTalk 2

(Semtation, 2005) uses Microsoft Visio’s

functionality to create and modify ontologies

graphically, again in a UML format, translating

ontologies automatically to an OWL ontology file.

Although ontology creation is aided by the

graphical functions of these tools they still remain

relatively complex, require a certain amount of

initial training; and are generic in nature and not

KMIS 2010 - International Conference on Knowledge Management and Information Sharing

208

designed specifically for security ontology creation

unlike our proposed tool.

The development of our tool has been based on

looking at the positive and relevant features from

currently available applications. The result is an

ontology editing tool designed specifically for

security domain experts, allowing them to capture

their knowledge in an easy and intuitive way while

removing the need to know of ontology construction

techniques.

2.4 Requirements

After review of related work and the problem space,

the main requirements of an ontology editing tool

for CISOs and HFRs have been identified:

 Encourage Knowledge Capture. The tool must

capture the unstructured knowledge of domain

experts within a suitable formalised ontology

structure.

 Hide Technical Details. A domain expert should

only need to concern themselves with adding

new information, without a need to manage the

underlying ontology structure.

 Assist Knowledge Owners. Knowledge owners

perceive their knowledge in their own way. To

formalise knowledge correctly, users should be

actively assisted by the editor tool during

knowledge entry

 Minimise Inconsistencies. There must be

mechanisms to minimise errors occurring in the

knowledge capture process.

3 IMPLEMENTATION

3.1 Components

In Figure 1 the ontology editor’s main components

can be seen.



Figure 1: Overview of ontology editor’s components.

3.1.1 Foundation Ontology

The editor uses an existing ontology (Parkin, 2009)

to define security issues and human-behavioural

factors, and crucially relate them to concepts within

an organisation’s security policies. For these

purposes we use a modified version of the ontology.

The concepts represented in the ontology are

shown in Figure 2. The ontology represents those

information

Assets that either must be secured or

which are crucial to an information security

management process. A Vulnerability may

represent the security and usability weaknesses of an

Asset that may promote or inhibit certain employee

behaviours. A

Vulnerability may be

‘exploitedBy’ a Threat which renders the

Figure 2: Overview of the information security and human

factors ontology.

Asset insecure or unusable (thereby also

potentially affecting productivity). Exploitation of a

Vulnerability may be intentional or accidental.

Threat may be either an Infrastructure

Threat or a Procedural Threat. The former

represent activities that directly affect security

mechanisms, whereas the latter represent security

events that impact upon an individual and their

behaviour. The

Behavioural Foundation of a

Procedural Threat classifies behaviours to

indicate the concerns that they raise within an

organisation (e.g., a person’s memory capabilities or

attitude towards security).

A Vulnerability may be ‘mitigatedBy’ a

Behaviour Control, which represents a

procedural activity that a CISO can enact to manage

interactions between humans and organisational

security controls. The associated

Risk Control

Type indicates a risk management approach, such

ONTOLOGY EDITING TOOL FOR INFORMATION SECURITY AND HUMAN FACTORS EXPERTS

209

that a

Behaviour Control ‘managesRiskOf’ a

Threat.

3.1.2 Ontology Editor

The ontology editor offers the user a simple

graphical interface where they can enter and capture

their knowledge in graphical form (i.e. a graphical

representation of an information security ontology).

All aspects of the security ontology structure

(e.g. classes and properties) are pre-defined, and

with the integrated help system, diagram

construction (i.e. ontology population) is intended to

be simplified and intuitive.

The underlying ontology is not presented to the

user but used implicitly to derive the graphical

elements and to translate the diagram into OWL.

Figure 3: Screenshot of editor’s user interface.

To the right of the user interface are the available

controls (new/open/save diagram, create OWL file,

etc).

The main window of the editor interface, as

shown in Figure 3, is where ontology diagram

construction takes place, using an embedded

instance of the Microsoft Visio 2007’s drawing

control (Microsoft, 2010). Use of Visio is

appropriate for our tool as it is intended for use in

large organisations that are assumed to already have

Microsoft Office software available.

On the left of the drawing control is a list of pre-

defined shapes available for constructing diagrams.

Each shape contains individual data (name, colour,

size, etc) and can be re-used by dragging and

dropping onto the drawing page. The shapes consist

of boxes (ontology concepts) and arrows (ontology

relationships), each represented in a different colour

to help the user differentiate between them.

When using arrows to form ontology

relationships, the Visio auto-connect feature is

applied to assist in the process. This same feature

also maintains connections between shapes when a

shape is moved around the drawing area. This then

allows the user to manipulate the diagram according

to their needs without needing to manage

connections.

Figure 4: Dialog box for adding new concept.

A number of mechanisms are used to restrict the

potential for errors in the knowledge capture

process. Only certain boxes can be associated with

each other using certain types of arrows, removing

the potential for invalid connections. A further error

handling feature of the editor is the detection of

unconnected shapes (isolated boxes, unconnected

arrows, etc). If such “hanging” shapes are found,

saving of the ontology file is halted until the user

resolves the relevant errors.

An integrated help system is in place to aid the

user in diagram construction. When a new box

(concept) is created a dialog box explains what the

box represents, how it is used and how it may be

connected to other boxes in the diagram. An

example is shown in Figure 4. “Tool tips” actively

provide explanations of ontology components and

editor functions when the user floats the mouse

cursor over them.

3.1.3 Ontology File

When an ontology diagram is saved, information

about the ontology content is stored in XML format.

This facilitates translation into an Ontology File,

wherein ontology content is translated into the Web

Ontology Language (OWL) (W3C, 2004) format.

KMIS 2010 - International Conference on Knowledge Management and Information Sharing

210



Figure 5: Overview of Translation Program’s components.

OWL files can potentially be processed

automatically by other software programs, thereby

providing scope for expert knowledge to be used in

various ways. Once created, all ontology files are

stored in a user designated file store.

Further knowledge may be added to pre-existing

ontology diagrams which is then propagated to the

corresponding ontology file. This removes the need

to regenerate the entire ontology.

3.1.4 Translation Program

The Translation Program processes Ontology

Diagrams to create Ontology Files.

The Java Translator Program is written in the

Java SE v1.5 programming language (Oracle, 2010)

and is deployed as an executable Java archive on a

user’s machine.

Figure 5 provides an overview of the

components involved in the translation process. This

process is two-fold: first an Ontology Diagram file is

processed to obtain relevant data; secondly, that data

is transformed into the OWL format before being

compiled into an ontology file. An entire diagram

can be translated into an ontology file via a single

operation, removing the need for successive

processing of each knowledge fragment.

Data is retrieved for each shape (instance) based

on its type, contained text (entered by user) and any

connections to other shapes. To write this data into

an ontology file, libraries from the OWL Java API

(Manchester, 2010) are used.

The classes and properties definitions of the

underlying information security ontology are hard-

coded into the Translation Program, and are written

directly to the ontology file. By predefining the

ontology structure the user is not expected to

understand or define ontology language constructs.

4 CONCLUSIONS

Information security management knowledge and

security human factors knowledge is currently vast

and fractured. This knowledge must be collated to

enable organisations to make more well informed

security policy decisions. Ontologies serve as a

means to store this knowledge.

Current ontology development utilises editing tools

which are mostly aimed at ontology experts, and not

necessarily those who hold the knowledge that must

be captured. This paper shows that there is a need

for an editing tool designed for information security

and human factors experts (amongst others) to

capture their interrelated knowledge in the form of

an ontology.

We describe a tool that produces machine-

readable OWL ontology files, derived from a

diagrammatic representation of information security

and human factors knowledge. Ontology concepts

and relationships are automatically translated from

diagram components into an OWL ontology file.

Our tool simplifies the ontology development

process, requiring little or no knowledge of ontology

construction on the part of the knowledge owner.

Future work hopes to see the development of a

community Web-based version of the editing tool to

facilitate collaborative knowledge capture across

wider communities of information security and

human factors experts.

ACKNOWLEDGEMENTS

The authors are supported in part by UK Technology

Strategy Board (TSB), grant nr. P0007E (“Trust

Economics”) and HP Labs Innovation Research

Program, award ID 2009-1052-1-A (“Prediction and

Provenance for Multi-Objective Information

Security Management”).

REFERENCES

Altova, 2010. SemanticWorks Semantic Web tool - Visual

RDF and OWL editor,

http://www.altova.com/semanticworks.html, last

visited 18/03/10

British Standards Institution (BSI UK), 2005. BS ISO/IEC

27001:2005 - Information Technology - Security

Techniques - Information Security Management

Systems – Requirements. BSI.

Fenz, S., Goluch, G. , Ekelhart, A. , Riedl, B., Weippl, E.,

2007. Information Security Fortification by

ONTOLOGY EDITING TOOL FOR INFORMATION SECURITY AND HUMAN FACTORS EXPERTS

211

Ontological Mapping of the ISO/IEC 270001

Standard. Proceedings of the 13

Pacific Rim

International Symposium on Dependable Computing

(PRDC2007). IEEE Computer Society.

Manchester, University of , 2010. The OWL API.

http://owlapi.sourceforge.net/, last visited 18/03/10

Microsoft Corporation, “Visio Home Page – Microsoft

Office Online”, 2010. http://office.microsoft.com/en-

us/visio/default.aspx, last viewed 18/03/10

MINDSWAP Reasearch Group, 2004. SWOOP – A

Hypermedia-based Featherweight OWL Ontology

Editor. http://www.mindswap.org/

2004/SWOOP/, last visited 22/07/10

NeOn, 2010. NeOn toolkit. http://neon-

toolkit.org/wiki/Main_Page, last visited 22/07/10

Ontoprise GmbH, 2010. OntoStudio.

http://www.ontoprise.de/en/home/products/

ontostudio/, last visited 22/07/10

Oracle Corporation, 2010. Java SE Overview - at a

Glance. http://java.sun.com/javase/, last visited

18/03/10

Parkin, S. E., van Moorsel, A., Coles. R., 2009. An

Information Security Ontology Incorporating Human-

Behavioural Implications, Proceedings of the 2

International Conference on Security of Information

and Networks (SIN).

Scicluna, J., Abela, C., Montebello, M., 2004. Visual

Modeling of OWL-S Services. Computer Science

Annual Research Workshop (CSAW) ’04.

Semtation GmbH, 2005. SemTalk – Ontology Editor for

MS Office.

http://www.semtalk.com/fr_index.html?/Ontologien.ht

m, last visited 18/03/10

Skidmore, P. “Beyond Measure”, Demos, 2003

Stanford Center for Biomedical Informatics Research,

2010. Welcome to Protégé.

http://protege.stanford.edu/, last viewed 18/03/10

TopQuadrant, 2010. TopBraid Composer,

http://www.topquadrant.com/products/TB_Composer.

html, last visited 23/07/10

Vermont, the University of, 2006. Welcome to GrOWL

Website, http://www.uvm.edu/~skrivov/growl/, last

viewed 18/03/10

W3C, 2004. OWL Web Ontology Language Overview.

http://www.w3.org/TR/owl-features/, last viewed

18/03/10

KMIS 2010 - International Conference on Knowledge Management and Information Sharing

212